區塊101
163 Posts
區塊101
Report
Block User
加密貨幣、Web3 與全球投資｜Bitcoin・Crypto・市場分析・投資觀察
幣安推薦邀請碼 BNLOVER
7 Following
99 Followers
48 Liked
Posts
All
Quotes
Videos
·
--
Jensen Huang has formally admitted to an extramarital affair, but the story has unexpectedly led to a company that purchased at least $4.8 billion worth of AI equipment—and an investigation by U.S. intelligence agencies!
On November 8, NVIDIA founder Jensen Huang publicly apologized to his wife, children, and employees, admitting that he had been unfaithful to his wife. On the same day, NVIDIA’s board announced that it had launched an independent investigation.
Who exactly is Alice Huang (Huang Le), the woman involved in the affair?
According to U.S. media reports, she is a Chinese businesswoman who previously held a senior position at Megaspeed, a Singapore-based AI computing company. Her publicly available résumé spans media, banking, securities, and investment, but some of her earlier experience has not been independently verified.
What really deserves attention is Megaspeed.
Reports say that from June 2024 to February 2026, companies affiliated with Megaspeed purchased at least $4.8 billion worth of AI equipment incorporating NVIDIA technology.
U.S. regulators had previously investigated whether Megaspeed was involved in circumventing restrictions on exports of advanced chips to China.
More sensitively, U.S. intelligence agencies reportedly suspect that Alice Huang is linked to Chinese intelligence activities and may have tried to use her relationship with Jensen Huang to learn about the AI chip business and influence U.S. export policy.
However, these remain allegations that have not been established in court.
The Singapore government has denied investigating her alleged intelligence connections, and NVIDIA’s board has said it currently has no reason to believe that company secrets or U.S. national security have been compromised.
The chairs of three U.S. House committees have asked Jensen Huang to attend a closed-door briefing on November 18.
What truly deserves attention in this case is no longer just Jensen Huang’s private relationship, but how NVIDIA manages major overseas customers, billions of dollars in AI equipment transactions, and potential conflicts of interest.
There is currently no public evidence that Jensen Huang disclosed confidential information or took part in intelligence activities.
Jensen Huang has formally admitted to an extramarital affair, but the story has unexpectedly led to a company that purchased at least $4.8 billion worth of AI equipment—and an investigation by U.S. intelligence agencies!
On November 8, NVIDIA founder Jensen Huang publicly apologized to his wife, children, and employees, admitting that he had been unfaithful to his wife. On the same day, NVIDIA’s board announced that it had launched an independent investigation.
Who exactly is Alice Huang (Huang Le), the woman involved in the affair?
According to U.S. media reports, she is a Chinese businesswoman who previously held a senior position at Megaspeed, a Singapore-based AI computing company. Her publicly available résumé spans media, banking, securities, and investment, but some of her earlier experience has not been independently verified.
What really deserves attention is Megaspeed.
Reports say that from June 2024 to February 2026, companies affiliated with Megaspeed purchased at least $4.8 billion worth of AI equipment incorporating NVIDIA technology.
U.S. regulators had previously investigated whether Megaspeed was involved in circumventing restrictions on exports of advanced chips to China.
More sensitively, U.S. intelligence agencies reportedly suspect that Alice Huang is linked to Chinese intelligence activities and may have tried to use her relationship with Jensen Huang to learn about the AI chip business and influence U.S. export policy.
However, these remain allegations that have not been established in court.
The Singapore government has denied investigating her alleged intelligence connections, and NVIDIA’s board has said it currently has no reason to believe that company secrets or U.S. national security have been compromised.
The chairs of three U.S. House committees have asked Jensen Huang to attend a closed-door briefing on November 18.
What truly deserves attention in this case is no longer just Jensen Huang’s private relationship, but how NVIDIA manages major overseas customers, billions of dollars in AI equipment transactions, and potential conflicts of interest.
There is currently no public evidence that Jensen Huang disclosed confidential information or took part in intelligence activities.
On November 8, NVIDIA founder Jensen Huang publicly apologized to his wife, children, and employees, admitting that he had been unfaithful to his wife. On the same day, NVIDIA’s board announced that it had launched an independent investigation.
Who exactly is Alice Huang (Huang Le), the woman involved in the affair?
According to U.S. media reports, she is a Chinese businesswoman who previously held a senior position at Megaspeed, a Singapore-based AI computing company. Her publicly available résumé spans media, banking, securities, and investment, but some of her earlier experience has not been independently verified.
What really deserves attention is Megaspeed.
Reports say that from June 2024 to February 2026, companies affiliated with Megaspeed purchased at least $4.8 billion worth of AI equipment incorporating NVIDIA technology.
U.S. regulators had previously investigated whether Megaspeed was involved in circumventing restrictions on exports of advanced chips to China.
More sensitively, U.S. intelligence agencies reportedly suspect that Alice Huang is linked to Chinese intelligence activities and may have tried to use her relationship with Jensen Huang to learn about the AI chip business and influence U.S. export policy.
However, these remain allegations that have not been established in court.
The Singapore government has denied investigating her alleged intelligence connections, and NVIDIA’s board has said it currently has no reason to believe that company secrets or U.S. national security have been compromised.
The chairs of three U.S. House committees have asked Jensen Huang to attend a closed-door briefing on November 18.
What truly deserves attention in this case is no longer just Jensen Huang’s private relationship, but how NVIDIA manages major overseas customers, billions of dollars in AI equipment transactions, and potential conflicts of interest.
There is currently no public evidence that Jensen Huang disclosed confidential information or took part in intelligence activities.
44
·
--
What is Alice Huang (Huang Le)ʼs background in the Jensen Huang affair scandal?
She is a Chinese businesswoman who previously held a senior position at Megaspeed, a Singaporean AI computing-power company.
But her publicly listed career history is quite unusual:
• She has claimed to have worked as a reporter for Hunan TV and China Central Television
• She worked in private banking at HSBC
• She worked at CITIC Securities
• She held a senior position at Shanghai Hexi Investment Management
• She later joined Singapore-based Megaspeed, entering the AI chip industry
However, U.S. media reports say some of her media and finance industry experience has not been verified by her former employers.
Megaspeed is particularly noteworthy.
The company reportedly purchased more than $4.8 billion worth of NVIDIA chips over two years and has also come under investigation over possible involvement in circumventing U.S. chip export restrictions.
U.S. intelligence agencies reportedly suspect Alice Huang of links to Chinese intelligence activities and are scrutinizing her close relationship with Jensen Huang.
On November 8, Jensen Huang admitted to being unfaithful during his marriage, and the NVIDIA board announced that it had launched an independent investigation.
But it is important to note that the Singaporean government has denied conducting any related espionage investigation into her, and NVIDIA has said there is currently no reason to believe that company secrets or national security have been compromised.
No public court ruling has found Alice Huang to be a spy.
What is truly worth investigating is how a businesswoman whose career spans media, finance, and AI computing power came to establish a business relationship with the world’s most important AI chip company.
What is Alice Huang (Huang Le)ʼs background in the Jensen Huang affair scandal?
She is a Chinese businesswoman who previously held a senior position at Megaspeed, a Singaporean AI computing-power company.
But her publicly listed career history is quite unusual:
• She has claimed to have worked as a reporter for Hunan TV and China Central Television
• She worked in private banking at HSBC
• She worked at CITIC Securities
• She held a senior position at Shanghai Hexi Investment Management
• She later joined Singapore-based Megaspeed, entering the AI chip industry
However, U.S. media reports say some of her media and finance industry experience has not been verified by her former employers.
Megaspeed is particularly noteworthy.
The company reportedly purchased more than $4.8 billion worth of NVIDIA chips over two years and has also come under investigation over possible involvement in circumventing U.S. chip export restrictions.
U.S. intelligence agencies reportedly suspect Alice Huang of links to Chinese intelligence activities and are scrutinizing her close relationship with Jensen Huang.
On November 8, Jensen Huang admitted to being unfaithful during his marriage, and the NVIDIA board announced that it had launched an independent investigation.
But it is important to note that the Singaporean government has denied conducting any related espionage investigation into her, and NVIDIA has said there is currently no reason to believe that company secrets or national security have been compromised.
No public court ruling has found Alice Huang to be a spy.
What is truly worth investigating is how a businesswoman whose career spans media, finance, and AI computing power came to establish a business relationship with the world’s most important AI chip company.
She is a Chinese businesswoman who previously held a senior position at Megaspeed, a Singaporean AI computing-power company.
But her publicly listed career history is quite unusual:
• She has claimed to have worked as a reporter for Hunan TV and China Central Television
• She worked in private banking at HSBC
• She worked at CITIC Securities
• She held a senior position at Shanghai Hexi Investment Management
• She later joined Singapore-based Megaspeed, entering the AI chip industry
However, U.S. media reports say some of her media and finance industry experience has not been verified by her former employers.
Megaspeed is particularly noteworthy.
The company reportedly purchased more than $4.8 billion worth of NVIDIA chips over two years and has also come under investigation over possible involvement in circumventing U.S. chip export restrictions.
U.S. intelligence agencies reportedly suspect Alice Huang of links to Chinese intelligence activities and are scrutinizing her close relationship with Jensen Huang.
On November 8, Jensen Huang admitted to being unfaithful during his marriage, and the NVIDIA board announced that it had launched an independent investigation.
But it is important to note that the Singaporean government has denied conducting any related espionage investigation into her, and NVIDIA has said there is currently no reason to believe that company secrets or national security have been compromised.
No public court ruling has found Alice Huang to be a spy.
What is truly worth investigating is how a businesswoman whose career spans media, finance, and AI computing power came to establish a business relationship with the world’s most important AI chip company.
55
·
--
Jensen Huang personally emailed all NVIDIA employees, admitting that he had been unfaithful to his wife!
He acknowledged that his actions had hurt his wife and children, that there was no excuse for them, and that he was willing to take full responsibility.
But this is about more than just his private life.
Because media reports related to the matter raised concerns about foreign intelligence activity, NVIDIA’s board has launched an independent investigation.
The company emphasized that there is currently no evidence that national security or NVIDIA’s business secrets have been compromised.
Huang said he would fully cooperate with the investigation, continue serving as CEO, and hoped the public would respect his family’s privacy.
Jensen Huang personally emailed all NVIDIA employees, admitting that he had been unfaithful to his wife!
He acknowledged that his actions had hurt his wife and children, that there was no excuse for them, and that he was willing to take full responsibility.
But this is about more than just his private life.
Because media reports related to the matter raised concerns about foreign intelligence activity, NVIDIA’s board has launched an independent investigation.
The company emphasized that there is currently no evidence that national security or NVIDIA’s business secrets have been compromised.
Huang said he would fully cooperate with the investigation, continue serving as CEO, and hoped the public would respect his family’s privacy.
He acknowledged that his actions had hurt his wife and children, that there was no excuse for them, and that he was willing to take full responsibility.
But this is about more than just his private life.
Because media reports related to the matter raised concerns about foreign intelligence activity, NVIDIA’s board has launched an independent investigation.
The company emphasized that there is currently no evidence that national security or NVIDIA’s business secrets have been compromised.
Huang said he would fully cooperate with the investigation, continue serving as CEO, and hoped the public would respect his family’s privacy.
73
·
--
Could Earth’s power supply already be insufficient for AI?
Elon Musk has proposed a very bold plan: sending AI computers into space.
According to him, SpaceX is preparing to co-design a space-based AI computer called StarMind with NVIDIA, with launches potentially beginning as early as the third quarter of 2027.
Even more surprising, this isn’t starting entirely from scratch.
Musk said SpaceX has already tested satellites equipped with NVIDIA GPUs, and recently launched a satellite carrying a Google TPU. The hardware has been operating well in space.
But why does he think AI needs to go into space?
The answer is electricity.
Musk gave a startling example:
Suppose AI computing power increases electricity demand by about 200 GW each year in the future. The U.S. power grid might struggle to handle it.
For comparison, the average power consumption in the United States is about 500 GW, while peak demand is around 1.2 TW.
In other words, if AI adds a sustained 200 GW of power load each year, that would be equivalent to adding about 40% of current average U.S. electricity consumption annually.
This is only a hypothetical scenario proposed by Musk; it does not mean AI is already consuming this much additional electricity every year.
He expects StarMind’s initial launches to begin in the third quarter of 2027, with deployments gradually increasing in 2028 and potentially reaching a much larger scale in 2029.
However, space-based AI computers still face many challenges.
These include how to dissipate heat, how to transmit large amounts of data back to Earth, satellite lifespans, and the enormous costs of launches and maintenance.
These problems haven’t all been solved simply because GPUs or TPUs can operate in orbit.
But this direction reflects an important trend:
The core of future AI competition may not just be who has the best models or the most chips, but who can secure enough energy to keep computing power running continuously.
If space-based data centers become a reality, the AI infrastructure industry could see a completely different competitive landscape.
#Block101 #SpaceX #NVIDIA #ElonMusk #AI #SpaceTechnology #ArtificialIntelligence
Could Earth’s power supply already be insufficient for AI?
Elon Musk has proposed a very bold plan: sending AI computers into space.
According to him, SpaceX is preparing to co-design a space-based AI computer called StarMind with NVIDIA, with launches potentially beginning as early as the third quarter of 2027.
Even more surprising, this isn’t starting entirely from scratch.
Musk said SpaceX has already tested satellites equipped with NVIDIA GPUs, and recently launched a satellite carrying a Google TPU. The hardware has been operating well in space.
But why does he think AI needs to go into space?
The answer is electricity.
Musk gave a startling example:
Suppose AI computing power increases electricity demand by about 200 GW each year in the future. The U.S. power grid might struggle to handle it.
For comparison, the average power consumption in the United States is about 500 GW, while peak demand is around 1.2 TW.
In other words, if AI adds a sustained 200 GW of power load each year, that would be equivalent to adding about 40% of current average U.S. electricity consumption annually.
This is only a hypothetical scenario proposed by Musk; it does not mean AI is already consuming this much additional electricity every year.
He expects StarMind’s initial launches to begin in the third quarter of 2027, with deployments gradually increasing in 2028 and potentially reaching a much larger scale in 2029.
However, space-based AI computers still face many challenges.
These include how to dissipate heat, how to transmit large amounts of data back to Earth, satellite lifespans, and the enormous costs of launches and maintenance.
These problems haven’t all been solved simply because GPUs or TPUs can operate in orbit.
But this direction reflects an important trend:
The core of future AI competition may not just be who has the best models or the most chips, but who can secure enough energy to keep computing power running continuously.
If space-based data centers become a reality, the AI infrastructure industry could see a completely different competitive landscape.
#Block101 #SpaceX #NVIDIA #ElonMusk #AI #SpaceTechnology #ArtificialIntelligence
Elon Musk has proposed a very bold plan: sending AI computers into space.
According to him, SpaceX is preparing to co-design a space-based AI computer called StarMind with NVIDIA, with launches potentially beginning as early as the third quarter of 2027.
Even more surprising, this isn’t starting entirely from scratch.
Musk said SpaceX has already tested satellites equipped with NVIDIA GPUs, and recently launched a satellite carrying a Google TPU. The hardware has been operating well in space.
But why does he think AI needs to go into space?
The answer is electricity.
Musk gave a startling example:
Suppose AI computing power increases electricity demand by about 200 GW each year in the future. The U.S. power grid might struggle to handle it.
For comparison, the average power consumption in the United States is about 500 GW, while peak demand is around 1.2 TW.
In other words, if AI adds a sustained 200 GW of power load each year, that would be equivalent to adding about 40% of current average U.S. electricity consumption annually.
This is only a hypothetical scenario proposed by Musk; it does not mean AI is already consuming this much additional electricity every year.
He expects StarMind’s initial launches to begin in the third quarter of 2027, with deployments gradually increasing in 2028 and potentially reaching a much larger scale in 2029.
However, space-based AI computers still face many challenges.
These include how to dissipate heat, how to transmit large amounts of data back to Earth, satellite lifespans, and the enormous costs of launches and maintenance.
These problems haven’t all been solved simply because GPUs or TPUs can operate in orbit.
But this direction reflects an important trend:
The core of future AI competition may not just be who has the best models or the most chips, but who can secure enough energy to keep computing power running continuously.
If space-based data centers become a reality, the AI infrastructure industry could see a completely different competitive landscape.
#Block101 #SpaceX #NVIDIA #ElonMusk #AI #SpaceTechnology #ArtificialIntelligence
296
·
--
The U.S. issued an order restricting fund transfers involving MEXC.
On November 6, the U.S. Financial Crimes Enforcement Network (FinCEN) issued an order prohibiting regulated U.S. financial institutions from transferring funds to MEXC or receiving funds from MEXC.
The restriction will take effect 30 days after the order is published in the Federal Register. The exact effective date depends on the official publication date and cannot be calculated simply by counting 30 days from the announcement date.
MEXC said that the platform is operating normally, user assets are safe, and an independent review has been initiated.
Note that the order applies to fund transfers between regulated U.S. financial institutions and MEXC; it should not be described as a blanket suspension of withdrawals by MEXC.
The U.S. issued an order restricting fund transfers involving MEXC.
On November 6, the U.S. Financial Crimes Enforcement Network (FinCEN) issued an order prohibiting regulated U.S. financial institutions from transferring funds to MEXC or receiving funds from MEXC.
The restriction will take effect 30 days after the order is published in the Federal Register. The exact effective date depends on the official publication date and cannot be calculated simply by counting 30 days from the announcement date.
MEXC said that the platform is operating normally, user assets are safe, and an independent review has been initiated.
Note that the order applies to fund transfers between regulated U.S. financial institutions and MEXC; it should not be described as a blanket suspension of withdrawals by MEXC.
On November 6, the U.S. Financial Crimes Enforcement Network (FinCEN) issued an order prohibiting regulated U.S. financial institutions from transferring funds to MEXC or receiving funds from MEXC.
The restriction will take effect 30 days after the order is published in the Federal Register. The exact effective date depends on the official publication date and cannot be calculated simply by counting 30 days from the announcement date.
MEXC said that the platform is operating normally, user assets are safe, and an independent review has been initiated.
Note that the order applies to fund transfers between regulated U.S. financial institutions and MEXC; it should not be described as a blanket suspension of withdrawals by MEXC.
105
·
--
The U.S. Congress may be gearing up for a new wave of large-scale corporate investigations, with the Trump family’s cryptocurrency businesses also in the spotlight!
The Washington Sun reported that senior Democratic lawmakers are reviewing a 181-page proposal for corporate investigations.
The document, drafted by the think tank Groundwork Collaborative, covers tech giants, AI companies, energy firms, financial institutions, and World Liberty Financial, which is associated with the Trump family.
The proposal aims to enable a future Congress to subpoena corporate executives, demand internal documents, and investigate potential conflicts of interest between companies and the government.
Notably, on November 5, Democratic Senator Richard Blumenthal separately requested documents from several companies in the technology and cryptocurrency sectors.
This means some oversight efforts are already underway, but it does not mean the 181-page proposal has been adopted in full.
For the cryptocurrency market, what’s worth watching next is whether companies such as World Liberty Financial will face formal hearings, and whether the next Congress will expand oversight of the crypto industry.
The U.S. Congress may be gearing up for a new wave of large-scale corporate investigations, with the Trump family’s cryptocurrency businesses also in the spotlight!
The Washington Sun reported that senior Democratic lawmakers are reviewing a 181-page proposal for corporate investigations.
The document, drafted by the think tank Groundwork Collaborative, covers tech giants, AI companies, energy firms, financial institutions, and World Liberty Financial, which is associated with the Trump family.
The proposal aims to enable a future Congress to subpoena corporate executives, demand internal documents, and investigate potential conflicts of interest between companies and the government.
Notably, on November 5, Democratic Senator Richard Blumenthal separately requested documents from several companies in the technology and cryptocurrency sectors.
This means some oversight efforts are already underway, but it does not mean the 181-page proposal has been adopted in full.
For the cryptocurrency market, what’s worth watching next is whether companies such as World Liberty Financial will face formal hearings, and whether the next Congress will expand oversight of the crypto industry.
The Washington Sun reported that senior Democratic lawmakers are reviewing a 181-page proposal for corporate investigations.
The document, drafted by the think tank Groundwork Collaborative, covers tech giants, AI companies, energy firms, financial institutions, and World Liberty Financial, which is associated with the Trump family.
The proposal aims to enable a future Congress to subpoena corporate executives, demand internal documents, and investigate potential conflicts of interest between companies and the government.
Notably, on November 5, Democratic Senator Richard Blumenthal separately requested documents from several companies in the technology and cryptocurrency sectors.
This means some oversight efforts are already underway, but it does not mean the 181-page proposal has been adopted in full.
For the cryptocurrency market, what’s worth watching next is whether companies such as World Liberty Financial will face formal hearings, and whether the next Congress will expand oversight of the crypto industry.
27
·
--
Taiwan stocks post biggest intraday point drop in history! The FSC is urgently considering measures to stabilize the market.
On November 6, Taiwan’s stock market saw sharp volatility, plunging as much as 3,737.67 points intraday, or 8.99%.
Although the decline narrowed to 4.51% by the close, the market still faces considerable uncertainty.
Key figures:
① Biggest intraday drop: 3,737.67 points (-8.99%)
② Closing decline: 1,873.08 points (-4.51%)
③ Major weighted stocks, including TSMC, briefly hit their daily price limits.
④ The FSC is preparing to hold a meeting to discuss measures to stabilize the stock market.
Possible measures currently being considered include restricting short selling through stock borrowing and raising margin requirements for short selling, but no measures have been officially confirmed for implementation.
This volatility also reflects the global AI and semiconductor industries undergoing a repricing.
As U.S. AI stocks such as NVIDIA and AMD fell sharply, Taiwan—as a key part of the global semiconductor supply chain—was also affected by market sentiment.
Three things to watch next:
Whether U.S. tech stocks can stabilize, whether the Taiwanese government will introduce market-stabilization measures, and whether global semiconductor investment continues to flow out.
#TaiwanStocks #TSMC #AI #Semiconductors #USStocks
Taiwan stocks post biggest intraday point drop in history! The FSC is urgently considering measures to stabilize the market.
On November 6, Taiwan’s stock market saw sharp volatility, plunging as much as 3,737.67 points intraday, or 8.99%.
Although the decline narrowed to 4.51% by the close, the market still faces considerable uncertainty.
Key figures:
① Biggest intraday drop: 3,737.67 points (-8.99%)
② Closing decline: 1,873.08 points (-4.51%)
③ Major weighted stocks, including TSMC, briefly hit their daily price limits.
④ The FSC is preparing to hold a meeting to discuss measures to stabilize the stock market.
Possible measures currently being considered include restricting short selling through stock borrowing and raising margin requirements for short selling, but no measures have been officially confirmed for implementation.
This volatility also reflects the global AI and semiconductor industries undergoing a repricing.
As U.S. AI stocks such as NVIDIA and AMD fell sharply, Taiwan—as a key part of the global semiconductor supply chain—was also affected by market sentiment.
Three things to watch next:
Whether U.S. tech stocks can stabilize, whether the Taiwanese government will introduce market-stabilization measures, and whether global semiconductor investment continues to flow out.
#TaiwanStocks #TSMC #AI #Semiconductors #USStocks
On November 6, Taiwan’s stock market saw sharp volatility, plunging as much as 3,737.67 points intraday, or 8.99%.
Although the decline narrowed to 4.51% by the close, the market still faces considerable uncertainty.
Key figures:
① Biggest intraday drop: 3,737.67 points (-8.99%)
② Closing decline: 1,873.08 points (-4.51%)
③ Major weighted stocks, including TSMC, briefly hit their daily price limits.
④ The FSC is preparing to hold a meeting to discuss measures to stabilize the stock market.
Possible measures currently being considered include restricting short selling through stock borrowing and raising margin requirements for short selling, but no measures have been officially confirmed for implementation.
This volatility also reflects the global AI and semiconductor industries undergoing a repricing.
As U.S. AI stocks such as NVIDIA and AMD fell sharply, Taiwan—as a key part of the global semiconductor supply chain—was also affected by market sentiment.
Three things to watch next:
Whether U.S. tech stocks can stabilize, whether the Taiwanese government will introduce market-stabilization measures, and whether global semiconductor investment continues to flow out.
#TaiwanStocks #TSMC #AI #Semiconductors #USStocks
81
·
--
What does the U.S. election have to do with my wallet? 😭
I bought stocks to earn 3% and hedge against inflation, and now they’re down 30%.
What does the U.S. election have to do with my wallet? 😭
I bought stocks to earn 3% and hedge against inflation, and now they’re down 30%.
I bought stocks to earn 3% and hedge against inflation, and now they’re down 30%.
39
·
--
France’s traditional bank Crédit Mutuel Arkéa has reportedly suffered a major cybersecurity incident involving approximately €100 million!
The incident has drawn attention because the target was not a cryptocurrency exchange, but a regulated traditional financial institution.
According to reports currently available, the attackers appear to have conducted abnormal transactions through the bank’s systems, involving approximately €100 million.
The bank says customer assets were not affected, but it remains unclear how much of the funds in question can be recovered.
Another noteworthy detail is that some of the stolen funds may have been converted into cryptocurrency.
However, this does not mean cryptocurrency was the cause of the bank hack.
The core of the incident remains unauthorized activity within the bank’s systems and how the funds were subsequently transferred.
As links between traditional financial institutions and digital asset markets increase, financial crime investigations increasingly need to span different payment systems.
Key questions to follow in this incident include:
How did the attackers gain access to the bank’s systems? Can the €100 million involved be recovered? Did any funds flow to cryptocurrency trading platforms? How will the bank improve its internal security controls going forward?
The investigation is ongoing, and the full attack method and final losses have yet to be confirmed.
France’s traditional bank Crédit Mutuel Arkéa has reportedly suffered a major cybersecurity incident involving approximately €100 million!
The incident has drawn attention because the target was not a cryptocurrency exchange, but a regulated traditional financial institution.
According to reports currently available, the attackers appear to have conducted abnormal transactions through the bank’s systems, involving approximately €100 million.
The bank says customer assets were not affected, but it remains unclear how much of the funds in question can be recovered.
Another noteworthy detail is that some of the stolen funds may have been converted into cryptocurrency.
However, this does not mean cryptocurrency was the cause of the bank hack.
The core of the incident remains unauthorized activity within the bank’s systems and how the funds were subsequently transferred.
As links between traditional financial institutions and digital asset markets increase, financial crime investigations increasingly need to span different payment systems.
Key questions to follow in this incident include:
How did the attackers gain access to the bank’s systems? Can the €100 million involved be recovered? Did any funds flow to cryptocurrency trading platforms? How will the bank improve its internal security controls going forward?
The investigation is ongoing, and the full attack method and final losses have yet to be confirmed.
The incident has drawn attention because the target was not a cryptocurrency exchange, but a regulated traditional financial institution.
According to reports currently available, the attackers appear to have conducted abnormal transactions through the bank’s systems, involving approximately €100 million.
The bank says customer assets were not affected, but it remains unclear how much of the funds in question can be recovered.
Another noteworthy detail is that some of the stolen funds may have been converted into cryptocurrency.
However, this does not mean cryptocurrency was the cause of the bank hack.
The core of the incident remains unauthorized activity within the bank’s systems and how the funds were subsequently transferred.
As links between traditional financial institutions and digital asset markets increase, financial crime investigations increasingly need to span different payment systems.
Key questions to follow in this incident include:
How did the attackers gain access to the bank’s systems? Can the €100 million involved be recovered? Did any funds flow to cryptocurrency trading platforms? How will the bank improve its internal security controls going forward?
The investigation is ongoing, and the full attack method and final losses have yet to be confirmed.
101
·
--
OpenAI officially launches GPT-6 Cyber, ushering in a new phase of the AI cybersecurity race!
OpenAI announced the launch of GPT-6 Cyber, designed specifically for cybersecurity research.
The model was used in OpenAI’s latest Linux kernel vulnerability research, and cybersecurity teams at several major companies have now begun using it in practice.
These include:
• Mozilla
• Palo Alto Networks
• Calif
• Cisco
• OpenAI’s own cybersecurity team
Unlike general-purpose AI models, GPT-6 Cyber focuses on security research, helping professional teams analyze system weaknesses and improve the efficiency of vulnerability research.
OpenAI also emphasized that while the new model enhances cybersecurity capabilities, it also strengthens safety alignment measures.
GPT-6 Cyber is currently available only to select Daybreak Red customers and has not been released to the public at large.
This is particularly noteworthy for the cryptocurrency industry.
Exchanges, wallets, smart contracts, and blockchain infrastructure all face highly complex cyberattacks.
In the future, AI may be used not only for trading strategies and software development, but also become an important tool for preventing major cybersecurity incidents.
However, the risk of AI being maliciously exploited makes these same capabilities all the more concerning.
#OpenAI #Cybersecurity #Aİ #Crypto
OpenAI officially launches GPT-6 Cyber, ushering in a new phase of the AI cybersecurity race!
OpenAI announced the launch of GPT-6 Cyber, designed specifically for cybersecurity research.
The model was used in OpenAI’s latest Linux kernel vulnerability research, and cybersecurity teams at several major companies have now begun using it in practice.
These include:
• Mozilla
• Palo Alto Networks
• Calif
• Cisco
• OpenAI’s own cybersecurity team
Unlike general-purpose AI models, GPT-6 Cyber focuses on security research, helping professional teams analyze system weaknesses and improve the efficiency of vulnerability research.
OpenAI also emphasized that while the new model enhances cybersecurity capabilities, it also strengthens safety alignment measures.
GPT-6 Cyber is currently available only to select Daybreak Red customers and has not been released to the public at large.
This is particularly noteworthy for the cryptocurrency industry.
Exchanges, wallets, smart contracts, and blockchain infrastructure all face highly complex cyberattacks.
In the future, AI may be used not only for trading strategies and software development, but also become an important tool for preventing major cybersecurity incidents.
However, the risk of AI being maliciously exploited makes these same capabilities all the more concerning.
#OpenAI #Cybersecurity #Aİ #Crypto
OpenAI announced the launch of GPT-6 Cyber, designed specifically for cybersecurity research.
The model was used in OpenAI’s latest Linux kernel vulnerability research, and cybersecurity teams at several major companies have now begun using it in practice.
These include:
• Mozilla
• Palo Alto Networks
• Calif
• Cisco
• OpenAI’s own cybersecurity team
Unlike general-purpose AI models, GPT-6 Cyber focuses on security research, helping professional teams analyze system weaknesses and improve the efficiency of vulnerability research.
OpenAI also emphasized that while the new model enhances cybersecurity capabilities, it also strengthens safety alignment measures.
GPT-6 Cyber is currently available only to select Daybreak Red customers and has not been released to the public at large.
This is particularly noteworthy for the cryptocurrency industry.
Exchanges, wallets, smart contracts, and blockchain infrastructure all face highly complex cyberattacks.
In the future, AI may be used not only for trading strategies and software development, but also become an important tool for preventing major cybersecurity incidents.
However, the risk of AI being maliciously exploited makes these same capabilities all the more concerning.
#OpenAI #Cybersecurity #Aİ #Crypto
101
·
--
OpenAI launches its Images 2.6 image model, letting ChatGPT generate animated GIFs directly.
It’s available starting today on all ChatGPT and Codex subscription plans. Users can generate animations from text or uploaded images, adjust the speed, and edit individual frames. Photos of frequently used people, pets, or characters can also be saved as “reference images” for reuse later.
For developers, two new API models are available: Flare focuses on speed, generating an image in as little as about 6 seconds, while Sunburst focuses on image quality and consistency across multiple rounds of edits. OpenAI says both are 50% faster than the previous version. API prices are unchanged from the previous generation: $30 per 1 million tokens for image generation. The API also adds SVG vector export, so generated images can be used directly on websites, in apps, and in print.
OpenAI says users worldwide generate 4 billion images every week.
OpenAI launches its Images 2.6 image model, letting ChatGPT generate animated GIFs directly.
It’s available starting today on all ChatGPT and Codex subscription plans. Users can generate animations from text or uploaded images, adjust the speed, and edit individual frames. Photos of frequently used people, pets, or characters can also be saved as “reference images” for reuse later.
For developers, two new API models are available: Flare focuses on speed, generating an image in as little as about 6 seconds, while Sunburst focuses on image quality and consistency across multiple rounds of edits. OpenAI says both are 50% faster than the previous version. API prices are unchanged from the previous generation: $30 per 1 million tokens for image generation. The API also adds SVG vector export, so generated images can be used directly on websites, in apps, and in print.
OpenAI says users worldwide generate 4 billion images every week.
It’s available starting today on all ChatGPT and Codex subscription plans. Users can generate animations from text or uploaded images, adjust the speed, and edit individual frames. Photos of frequently used people, pets, or characters can also be saved as “reference images” for reuse later.
For developers, two new API models are available: Flare focuses on speed, generating an image in as little as about 6 seconds, while Sunburst focuses on image quality and consistency across multiple rounds of edits. OpenAI says both are 50% faster than the previous version. API prices are unchanged from the previous generation: $30 per 1 million tokens for image generation. The API also adds SVG vector export, so generated images can be used directly on websites, in apps, and in print.
OpenAI says users worldwide generate 4 billion images every week.
46
·
--
Cat Wu, former product lead for Claude Code, is joining OpenAI—a noteworthy personnel move in the AI development tools space.
Cat Wu recently left Anthropic and then announced she was joining OpenAI to lead product management for its core product team. Business Insider reported that OpenAI has confirmed the appointment.
Claude Code, which Cat Wu previously led, is Anthropic’s AI coding tool. Developers can assign it tasks and have it read projects, modify files, and run commands to help fix bugs, write tests, or develop new features. Users can then review the results.
OpenAI also has its own AI coding tool, Codex. Both companies want AI to take part in more software development work, and Cat Wu’s move from Anthropic to OpenAI has drawn attention to what she will work on next.
The role announced so far is product management for core products; it has not been announced that she will take over Codex.
3 | Facebook
Just after announcing her departure from Anthropic, Cat Wu, former product lead for Claude Code, revealed her next destination: OpenAI.
She will lead product management for OpenAI’s core product team, reporting to Tibo Sottiaux. Business Insider reported that OpenAI has confirmed the personnel move.
Cat Wu previously worked as an engineer and has experience in venture capital. She later led product work at Anthropic, including for Claude Code.
For those unfamiliar with it, Claude Code is an AI coding tool. Users can describe what they want to accomplish and have it read projects, modify code, and run commands to help develop features, fix bugs, or write tests.
For example, if a feature on a website isn’t working, a developer can ask it to review the relevant code, suggest changes, and run tests, then have a person review the results.
OpenAI also has its own Codex, and both companies are competing for developers. Now, Cat Wu, who once led the Claude Code product, is joining a direct competitor.
However, her new role is to lead product management for core products, a broader remit than any single coding tool. There has been no announcement that she will directly take over Codex.
Cat Wu, former product lead for Claude Code, is joining OpenAI—a noteworthy personnel move in the AI development tools space.
Cat Wu recently left Anthropic and then announced she was joining OpenAI to lead product management for its core product team. Business Insider reported that OpenAI has confirmed the appointment.
Claude Code, which Cat Wu previously led, is Anthropic’s AI coding tool. Developers can assign it tasks and have it read projects, modify files, and run commands to help fix bugs, write tests, or develop new features. Users can then review the results.
OpenAI also has its own AI coding tool, Codex. Both companies want AI to take part in more software development work, and Cat Wu’s move from Anthropic to OpenAI has drawn attention to what she will work on next.
The role announced so far is product management for core products; it has not been announced that she will take over Codex.
3 | Facebook
Just after announcing her departure from Anthropic, Cat Wu, former product lead for Claude Code, revealed her next destination: OpenAI.
She will lead product management for OpenAI’s core product team, reporting to Tibo Sottiaux. Business Insider reported that OpenAI has confirmed the personnel move.
Cat Wu previously worked as an engineer and has experience in venture capital. She later led product work at Anthropic, including for Claude Code.
For those unfamiliar with it, Claude Code is an AI coding tool. Users can describe what they want to accomplish and have it read projects, modify code, and run commands to help develop features, fix bugs, or write tests.
For example, if a feature on a website isn’t working, a developer can ask it to review the relevant code, suggest changes, and run tests, then have a person review the results.
OpenAI also has its own Codex, and both companies are competing for developers. Now, Cat Wu, who once led the Claude Code product, is joining a direct competitor.
However, her new role is to lead product management for core products, a broader remit than any single coding tool. There has been no announcement that she will directly take over Codex.
Cat Wu recently left Anthropic and then announced she was joining OpenAI to lead product management for its core product team. Business Insider reported that OpenAI has confirmed the appointment.
Claude Code, which Cat Wu previously led, is Anthropic’s AI coding tool. Developers can assign it tasks and have it read projects, modify files, and run commands to help fix bugs, write tests, or develop new features. Users can then review the results.
OpenAI also has its own AI coding tool, Codex. Both companies want AI to take part in more software development work, and Cat Wu’s move from Anthropic to OpenAI has drawn attention to what she will work on next.
The role announced so far is product management for core products; it has not been announced that she will take over Codex.
3 | Facebook
Just after announcing her departure from Anthropic, Cat Wu, former product lead for Claude Code, revealed her next destination: OpenAI.
She will lead product management for OpenAI’s core product team, reporting to Tibo Sottiaux. Business Insider reported that OpenAI has confirmed the personnel move.
Cat Wu previously worked as an engineer and has experience in venture capital. She later led product work at Anthropic, including for Claude Code.
For those unfamiliar with it, Claude Code is an AI coding tool. Users can describe what they want to accomplish and have it read projects, modify code, and run commands to help develop features, fix bugs, or write tests.
For example, if a feature on a website isn’t working, a developer can ask it to review the relevant code, suggest changes, and run tests, then have a person review the results.
OpenAI also has its own Codex, and both companies are competing for developers. Now, Cat Wu, who once led the Claude Code product, is joining a direct competitor.
However, her new role is to lead product management for core products, a broader remit than any single coding tool. There has been no announcement that she will directly take over Codex.
61
·
--
USDC issuer Circle posted a profit of about $226 million in its latest quarter, with revenue up 17% year over year—yet it also cut around 5% of its workforce!
There are several noteworthy details in this earnings report.
First, Circle remains profitable.
But the company’s revenue mix is changing.
According to the earnings report, about $182 million of this quarter’s revenue came from items related to the ARC token presale, while traditional reserve interest income fell by about 8%.
This suggests Circle is trying to reduce its reliance on interest income from stablecoin reserves.
In the past, a major source of Circle’s revenue was the interest earned on USDC reserve assets.
When the amount of USDC in circulation increases or market interest rates remain high, this business can usually generate substantial revenue.
But if interest rates fall, the yield on reserve assets may also be affected.
Circle is therefore developing more businesses related to blockchain infrastructure.
ARC is seen as one of the company’s key areas of focus for the future.
Another point worth noting is the layoffs.
Even though the company remains profitable, management has still decided to reduce its workforce by around 5%.
This also reflects how crypto companies are reassessing their cost structures and resource allocation.
Whether Circle can build more stable and diversified revenue streams in the future may matter more than its profit in any single quarter.
USDC issuer Circle posted a profit of about $226 million in its latest quarter, with revenue up 17% year over year—yet it also cut around 5% of its workforce!
There are several noteworthy details in this earnings report.
First, Circle remains profitable.
But the company’s revenue mix is changing.
According to the earnings report, about $182 million of this quarter’s revenue came from items related to the ARC token presale, while traditional reserve interest income fell by about 8%.
This suggests Circle is trying to reduce its reliance on interest income from stablecoin reserves.
In the past, a major source of Circle’s revenue was the interest earned on USDC reserve assets.
When the amount of USDC in circulation increases or market interest rates remain high, this business can usually generate substantial revenue.
But if interest rates fall, the yield on reserve assets may also be affected.
Circle is therefore developing more businesses related to blockchain infrastructure.
ARC is seen as one of the company’s key areas of focus for the future.
Another point worth noting is the layoffs.
Even though the company remains profitable, management has still decided to reduce its workforce by around 5%.
This also reflects how crypto companies are reassessing their cost structures and resource allocation.
Whether Circle can build more stable and diversified revenue streams in the future may matter more than its profit in any single quarter.
There are several noteworthy details in this earnings report.
First, Circle remains profitable.
But the company’s revenue mix is changing.
According to the earnings report, about $182 million of this quarter’s revenue came from items related to the ARC token presale, while traditional reserve interest income fell by about 8%.
This suggests Circle is trying to reduce its reliance on interest income from stablecoin reserves.
In the past, a major source of Circle’s revenue was the interest earned on USDC reserve assets.
When the amount of USDC in circulation increases or market interest rates remain high, this business can usually generate substantial revenue.
But if interest rates fall, the yield on reserve assets may also be affected.
Circle is therefore developing more businesses related to blockchain infrastructure.
ARC is seen as one of the company’s key areas of focus for the future.
Another point worth noting is the layoffs.
Even though the company remains profitable, management has still decided to reduce its workforce by around 5%.
This also reflects how crypto companies are reassessing their cost structures and resource allocation.
Whether Circle can build more stable and diversified revenue streams in the future may matter more than its profit in any single quarter.
1
245
·
--
Elon Musk spent about $446 million helping Republicans in the U.S. midterm elections, but Democrats won both chambers of Congress.
According to federal filings compiled by Newsweek, Musk was the biggest donor in this midterm election, contributing about three times as much as the runner-up. His biggest bet was the Texas Senate race: he put more than $200 million behind Republican Ken Paxton, but Democrat James Talarico won, becoming the first Democrat since 1988 to win a Senate seat in Texas. Greg Abbott, the Texas governor he supported with donations, also lost.
After the votes were counted, Musk joked on X: “Looks like my election-interfering space lasers didn’t work this time.” He added: “Space data centers are looking pretty tempting now.”
Tesla and SpaceX shares were both down more than 4% in premarket trading on Wednesday. Democrats had previously said that, if they took control of Congress, they would investigate Musk’s government business and federal contracts.
One of the key issues in the Texas campaign was AI data centers. Republicans want to turn Texas into a hub for the AI industry, while Democrats advocate stricter regulation and have called for data centers to bear the cost of their own infrastructure.
Elon Musk spent about $446 million helping Republicans in the U.S. midterm elections, but Democrats won both chambers of Congress.
According to federal filings compiled by Newsweek, Musk was the biggest donor in this midterm election, contributing about three times as much as the runner-up. His biggest bet was the Texas Senate race: he put more than $200 million behind Republican Ken Paxton, but Democrat James Talarico won, becoming the first Democrat since 1988 to win a Senate seat in Texas. Greg Abbott, the Texas governor he supported with donations, also lost.
After the votes were counted, Musk joked on X: “Looks like my election-interfering space lasers didn’t work this time.” He added: “Space data centers are looking pretty tempting now.”
Tesla and SpaceX shares were both down more than 4% in premarket trading on Wednesday. Democrats had previously said that, if they took control of Congress, they would investigate Musk’s government business and federal contracts.
One of the key issues in the Texas campaign was AI data centers. Republicans want to turn Texas into a hub for the AI industry, while Democrats advocate stricter regulation and have called for data centers to bear the cost of their own infrastructure.
According to federal filings compiled by Newsweek, Musk was the biggest donor in this midterm election, contributing about three times as much as the runner-up. His biggest bet was the Texas Senate race: he put more than $200 million behind Republican Ken Paxton, but Democrat James Talarico won, becoming the first Democrat since 1988 to win a Senate seat in Texas. Greg Abbott, the Texas governor he supported with donations, also lost.
After the votes were counted, Musk joked on X: “Looks like my election-interfering space lasers didn’t work this time.” He added: “Space data centers are looking pretty tempting now.”
Tesla and SpaceX shares were both down more than 4% in premarket trading on Wednesday. Democrats had previously said that, if they took control of Congress, they would investigate Musk’s government business and federal contracts.
One of the key issues in the Texas campaign was AI data centers. Republicans want to turn Texas into a hub for the AI industry, while Democrats advocate stricter regulation and have called for data centers to bear the cost of their own infrastructure.
1
355
·
--
Canva announced a major organizational restructuring, affecting 1,271 employees—about 22% of the company! But the most surprising part was the severance package: • At least six months’ salary • Employees can keep their company-issued MacBooks • An additional 12 months of health coverage in some regions • One year of LinkedIn Premium plus job-search support • Those who want to start a business can apply for a $10,000 grant What’s even more unusual is that Canva won’t ask for any equity, even if employees use the grant to start a new company. Founder Melanie Perkins admitted that the company’s growth has slowed for the first time in a decade, making this one of the most difficult moments in its 13-year history. However, the 1,271 people affected were not all laid off; the figure also includes role changes and relocations. The startup grant requires an application.
Canva announced a major organizational restructuring, affecting 1,271 employees—about 22% of the company! But the most surprising part was the severance package: • At least six months’ salary • Employees can keep their company-issued MacBooks • An additional 12 months of health coverage in some regions • One year of LinkedIn Premium plus job-search support • Those who want to start a business can apply for a $10,000 grant What’s even more unusual is that Canva won’t ask for any equity, even if employees use the grant to start a new company. Founder Melanie Perkins admitted that the company’s growth has slowed for the first time in a decade, making this one of the most difficult moments in its 13-year history. However, the 1,271 people affected were not all laid off; the figure also includes role changes and relocations. The startup grant requires an application.
118
·
--
North Korea Denies $6.4 Billion Binance Heist; Kim Yo Jong: “Let’s See Who Ends Up With the Recovered Money”
On November 3, North Korea’s state-run Korean Central News Agency published remarks by Kim Yo Jong denying any connection between North Korea and the Binance hack. She called the U.S. accusations “absurd fabrications, nothing new” and “a serious political provocation.”
Her argument was that the current U.S. administration announced as soon as it took office last year that it would add seized illicit crypto assets to the national strategic reserve. So, “if this so-called stolen money is recovered, perhaps we’ll find out who takes it all in the end—and who is behind it.” She ended by warning that North Korea’s hostility toward the U.S. “needs to be reconsidered in several respects.”
Timeline:
10/29, 20:05 to 23:56 UTC: Wallets on 11 Binance-supported chains were robbed of $6.43 billion (according to Elliptic; Binance puts the figure at about $6.3 billion). It was the largest crypto heist in history, worth more than four times the $1.5 billion Bybit theft in 2025. Binance said the incident began when an engineer’s laptop was compromised, that it found no evidence of private keys being leaked, and that it would reimburse users in full.
10/31: The FBI attributed the attack to TraderTraitor, a North Korean state-sponsored hacking group, and released the hackers’ addresses. TRM Labs and Chainalysis also attributed it to North Korea.
11/1: Binance resumed withdrawals. From 06:00 to 15:30 UTC, $4.649 billion was withdrawn and $1.01 billion deposited, for net outflows of about $3.64 billion.
11/2: North Korea issued Kim Yo Jong’s remarks; they were published on 11/3.
Tracking the funds (Elliptic):
・$2.85 billion has been frozen or seized, accounting for 44%. At the time of the Bybit hack, only about 3% had been frozen.
・Circle froze all $823 million in USDC. Tether froze the USDT 37 minutes after it entered Ethereum.
・About $786 million in stablecoins was exchanged before being frozen.
・The 18,270 stolen BTC are spread across 19 addresses and had not moved as of Elliptic’s latest update.
・CoinJoin mixing was first detected on 11/1.
・About $3.24 billion remains, mostly in ETH and BTC.
Elliptic expects the hackers to convert ETH to BTC next, then gradually cash out through over-the-counter (OTC) brokers and exchanges—a process that could take several months. In the Bybit case, North Korea laundered 499,000 stolen ETH in just 10 days, but cashing out took several months.
North Korea has denied all accusations of crypto theft in the past, including in the Bybit case. The U.S. has not yet responded to these remarks.
North Korea Denies $6.4 Billion Binance Heist; Kim Yo Jong: “Let’s See Who Ends Up With the Recovered Money”
On November 3, North Korea’s state-run Korean Central News Agency published remarks by Kim Yo Jong denying any connection between North Korea and the Binance hack. She called the U.S. accusations “absurd fabrications, nothing new” and “a serious political provocation.”
Her argument was that the current U.S. administration announced as soon as it took office last year that it would add seized illicit crypto assets to the national strategic reserve. So, “if this so-called stolen money is recovered, perhaps we’ll find out who takes it all in the end—and who is behind it.” She ended by warning that North Korea’s hostility toward the U.S. “needs to be reconsidered in several respects.”
Timeline:
10/29, 20:05 to 23:56 UTC: Wallets on 11 Binance-supported chains were robbed of $6.43 billion (according to Elliptic; Binance puts the figure at about $6.3 billion). It was the largest crypto heist in history, worth more than four times the $1.5 billion Bybit theft in 2025. Binance said the incident began when an engineer’s laptop was compromised, that it found no evidence of private keys being leaked, and that it would reimburse users in full.
10/31: The FBI attributed the attack to TraderTraitor, a North Korean state-sponsored hacking group, and released the hackers’ addresses. TRM Labs and Chainalysis also attributed it to North Korea.
11/1: Binance resumed withdrawals. From 06:00 to 15:30 UTC, $4.649 billion was withdrawn and $1.01 billion deposited, for net outflows of about $3.64 billion.
11/2: North Korea issued Kim Yo Jong’s remarks; they were published on 11/3.
Tracking the funds (Elliptic):
・$2.85 billion has been frozen or seized, accounting for 44%. At the time of the Bybit hack, only about 3% had been frozen.
・Circle froze all $823 million in USDC. Tether froze the USDT 37 minutes after it entered Ethereum.
・About $786 million in stablecoins was exchanged before being frozen.
・The 18,270 stolen BTC are spread across 19 addresses and had not moved as of Elliptic’s latest update.
・CoinJoin mixing was first detected on 11/1.
・About $3.24 billion remains, mostly in ETH and BTC.
Elliptic expects the hackers to convert ETH to BTC next, then gradually cash out through over-the-counter (OTC) brokers and exchanges—a process that could take several months. In the Bybit case, North Korea laundered 499,000 stolen ETH in just 10 days, but cashing out took several months.
North Korea has denied all accusations of crypto theft in the past, including in the Bybit case. The U.S. has not yet responded to these remarks.
On November 3, North Korea’s state-run Korean Central News Agency published remarks by Kim Yo Jong denying any connection between North Korea and the Binance hack. She called the U.S. accusations “absurd fabrications, nothing new” and “a serious political provocation.”
Her argument was that the current U.S. administration announced as soon as it took office last year that it would add seized illicit crypto assets to the national strategic reserve. So, “if this so-called stolen money is recovered, perhaps we’ll find out who takes it all in the end—and who is behind it.” She ended by warning that North Korea’s hostility toward the U.S. “needs to be reconsidered in several respects.”
Timeline:
10/29, 20:05 to 23:56 UTC: Wallets on 11 Binance-supported chains were robbed of $6.43 billion (according to Elliptic; Binance puts the figure at about $6.3 billion). It was the largest crypto heist in history, worth more than four times the $1.5 billion Bybit theft in 2025. Binance said the incident began when an engineer’s laptop was compromised, that it found no evidence of private keys being leaked, and that it would reimburse users in full.
10/31: The FBI attributed the attack to TraderTraitor, a North Korean state-sponsored hacking group, and released the hackers’ addresses. TRM Labs and Chainalysis also attributed it to North Korea.
11/1: Binance resumed withdrawals. From 06:00 to 15:30 UTC, $4.649 billion was withdrawn and $1.01 billion deposited, for net outflows of about $3.64 billion.
11/2: North Korea issued Kim Yo Jong’s remarks; they were published on 11/3.
Tracking the funds (Elliptic):
・$2.85 billion has been frozen or seized, accounting for 44%. At the time of the Bybit hack, only about 3% had been frozen.
・Circle froze all $823 million in USDC. Tether froze the USDT 37 minutes after it entered Ethereum.
・About $786 million in stablecoins was exchanged before being frozen.
・The 18,270 stolen BTC are spread across 19 addresses and had not moved as of Elliptic’s latest update.
・CoinJoin mixing was first detected on 11/1.
・About $3.24 billion remains, mostly in ETH and BTC.
Elliptic expects the hackers to convert ETH to BTC next, then gradually cash out through over-the-counter (OTC) brokers and exchanges—a process that could take several months. In the Bybit case, North Korea laundered 499,000 stolen ETH in just 10 days, but cashing out took several months.
North Korea has denied all accusations of crypto theft in the past, including in the Bybit case. The U.S. has not yet responded to these remarks.
143
·
--
$115 billion! A mega-merger is shaking up the AI optical communications industry.
Marvell (MRVL) has officially announced its acquisition of optical communications giant Lumentum (LITE), a deal that could reshape the competitive landscape for AI data center infrastructure.
Deal highlights:
① Enterprise value: $115 billion, including debt.
② Deal structure: All-stock exchange; each LITE share will be exchanged for 4.5 MRVL shares.
③ Acquisition premium: Approximately 23.9%.
④ Ownership structure: After the merger, existing Marvell shareholders will own 67%, and existing Lumentum shareholders will own 33%.
⑤ Expected completion: First half of 2028, subject to relevant approvals.
Why is this deal worth watching?
As AI models continue to grow, data centers need to transfer data at high speeds among large numbers of GPUs and servers.
Traditional copper connections face challenges in distance, bandwidth, and power consumption, increasing the importance of optical communications technology.
Marvell already has a presence in custom AI chips and high-speed networking technology. Acquiring Lumentum will further integrate its laser, optical transceiver, and optical switching capabilities.
This could also intensify competition between Marvell and Broadcom (AVGO) in the AI infrastructure market.
Could the next focus of AI investment shift from GPUs to optical communications?
#Aİ #Marvell #Lumentum #Semiconductor
$115 billion! A mega-merger is shaking up the AI optical communications industry.
Marvell (MRVL) has officially announced its acquisition of optical communications giant Lumentum (LITE), a deal that could reshape the competitive landscape for AI data center infrastructure.
Deal highlights:
① Enterprise value: $115 billion, including debt.
② Deal structure: All-stock exchange; each LITE share will be exchanged for 4.5 MRVL shares.
③ Acquisition premium: Approximately 23.9%.
④ Ownership structure: After the merger, existing Marvell shareholders will own 67%, and existing Lumentum shareholders will own 33%.
⑤ Expected completion: First half of 2028, subject to relevant approvals.
Why is this deal worth watching?
As AI models continue to grow, data centers need to transfer data at high speeds among large numbers of GPUs and servers.
Traditional copper connections face challenges in distance, bandwidth, and power consumption, increasing the importance of optical communications technology.
Marvell already has a presence in custom AI chips and high-speed networking technology. Acquiring Lumentum will further integrate its laser, optical transceiver, and optical switching capabilities.
This could also intensify competition between Marvell and Broadcom (AVGO) in the AI infrastructure market.
Could the next focus of AI investment shift from GPUs to optical communications?
#Aİ #Marvell #Lumentum #Semiconductor
Marvell (MRVL) has officially announced its acquisition of optical communications giant Lumentum (LITE), a deal that could reshape the competitive landscape for AI data center infrastructure.
Deal highlights:
① Enterprise value: $115 billion, including debt.
② Deal structure: All-stock exchange; each LITE share will be exchanged for 4.5 MRVL shares.
③ Acquisition premium: Approximately 23.9%.
④ Ownership structure: After the merger, existing Marvell shareholders will own 67%, and existing Lumentum shareholders will own 33%.
⑤ Expected completion: First half of 2028, subject to relevant approvals.
Why is this deal worth watching?
As AI models continue to grow, data centers need to transfer data at high speeds among large numbers of GPUs and servers.
Traditional copper connections face challenges in distance, bandwidth, and power consumption, increasing the importance of optical communications technology.
Marvell already has a presence in custom AI chips and high-speed networking technology. Acquiring Lumentum will further integrate its laser, optical transceiver, and optical switching capabilities.
This could also intensify competition between Marvell and Broadcom (AVGO) in the AI infrastructure market.
Could the next focus of AI investment shift from GPUs to optical communications?
#Aİ #Marvell #Lumentum #Semiconductor
45
·
--
Even Musk admits that operating AI agents could be prohibitively expensive!
On November 1, a paying Grok Bot user publicly complained that their weekly usage allowance ran out too quickly and asked the platform to loosen its limits.
Musk replied personally, saying that Grok Bot’s operating costs are currently too high relative to its revenue to be sustainable in the long run.
This brief exchange reflects a business challenge facing the entire AI agent industry.
Traditional chatbots mainly answer questions.
AI agents are different.
They may need to:
Carry out tasks involving multiple steps. Repeatedly call large language models. Search for information, use tools, and run programs. Retry after failures. Occupy computing resources for long periods.
This means that the actual cost of serving the same paying user can vary greatly depending on the complexity of their tasks.
If a platform charges a fixed monthly fee while allowing many costly operations, it must strike a balance between usage limits, model efficiency, and subscription prices.
Musk did not disclose Grok Bot’s actual losses, revenue, or computing costs, nor did he officially announce any price changes.
But the remarks are still worth paying attention to.
Competition among AI agents is no longer just about model capabilities.
The ability to complete more tasks at lower cost and build a sustainable pricing model could become a key competitive factor in the next phase.
Even Musk admits that operating AI agents could be prohibitively expensive!
On November 1, a paying Grok Bot user publicly complained that their weekly usage allowance ran out too quickly and asked the platform to loosen its limits.
Musk replied personally, saying that Grok Bot’s operating costs are currently too high relative to its revenue to be sustainable in the long run.
This brief exchange reflects a business challenge facing the entire AI agent industry.
Traditional chatbots mainly answer questions.
AI agents are different.
They may need to:
Carry out tasks involving multiple steps. Repeatedly call large language models. Search for information, use tools, and run programs. Retry after failures. Occupy computing resources for long periods.
This means that the actual cost of serving the same paying user can vary greatly depending on the complexity of their tasks.
If a platform charges a fixed monthly fee while allowing many costly operations, it must strike a balance between usage limits, model efficiency, and subscription prices.
Musk did not disclose Grok Bot’s actual losses, revenue, or computing costs, nor did he officially announce any price changes.
But the remarks are still worth paying attention to.
Competition among AI agents is no longer just about model capabilities.
The ability to complete more tasks at lower cost and build a sustainable pricing model could become a key competitive factor in the next phase.
On November 1, a paying Grok Bot user publicly complained that their weekly usage allowance ran out too quickly and asked the platform to loosen its limits.
Musk replied personally, saying that Grok Bot’s operating costs are currently too high relative to its revenue to be sustainable in the long run.
This brief exchange reflects a business challenge facing the entire AI agent industry.
Traditional chatbots mainly answer questions.
AI agents are different.
They may need to:
Carry out tasks involving multiple steps. Repeatedly call large language models. Search for information, use tools, and run programs. Retry after failures. Occupy computing resources for long periods.
This means that the actual cost of serving the same paying user can vary greatly depending on the complexity of their tasks.
If a platform charges a fixed monthly fee while allowing many costly operations, it must strike a balance between usage limits, model efficiency, and subscription prices.
Musk did not disclose Grok Bot’s actual losses, revenue, or computing costs, nor did he officially announce any price changes.
But the remarks are still worth paying attention to.
Competition among AI agents is no longer just about model capabilities.
The ability to complete more tasks at lower cost and build a sustainable pricing model could become a key competitive factor in the next phase.
87
·
--
After watching the charts and scrolling through news all day, did your judgment really improve?
In an interview, Joe Hudson, who coaches the OpenAI team, offered a thought-provoking perspective: the importance of hard work is overstated.
He believes that the quality of your decisions and leverage matter more than simply extending your work hours. He even suggested that making five or six excellent decisions a day could lead to significant results.
He has also observed that when Silicon Valley startups he knows fail, it’s often not because the team didn’t work hard enough.
This is his experience, not a formula for success, and it doesn’t guarantee investment returns. But in the information-overloaded crypto market, it’s worth considering: how much of the news you read today actually changed your judgment?
After watching the charts and scrolling through news all day, did your judgment really improve?
In an interview, Joe Hudson, who coaches the OpenAI team, offered a thought-provoking perspective: the importance of hard work is overstated.
He believes that the quality of your decisions and leverage matter more than simply extending your work hours. He even suggested that making five or six excellent decisions a day could lead to significant results.
He has also observed that when Silicon Valley startups he knows fail, it’s often not because the team didn’t work hard enough.
This is his experience, not a formula for success, and it doesn’t guarantee investment returns. But in the information-overloaded crypto market, it’s worth considering: how much of the news you read today actually changed your judgment?
In an interview, Joe Hudson, who coaches the OpenAI team, offered a thought-provoking perspective: the importance of hard work is overstated.
He believes that the quality of your decisions and leverage matter more than simply extending your work hours. He even suggested that making five or six excellent decisions a day could lead to significant results.
He has also observed that when Silicon Valley startups he knows fail, it’s often not because the team didn’t work hard enough.
This is his experience, not a formula for success, and it doesn’t guarantee investment returns. But in the information-overloaded crypto market, it’s worth considering: how much of the news you read today actually changed your judgment?
77
·
--
FBI officially attributes the theft of over $6 billion in assets from Binance to a North Korean hacking group!
On October 31, the U.S. Federal Bureau of Investigation (FBI) issued a notice stating that North Korean malicious cyber activity known as TraderTraitor was involved in the October 29 theft from Binance.
The announcement highlighted four major points:
① Official attribution: The FBI attributed the attack to North Korea’s TraderTraitor hacking activity.
② Fund transfers: Investigators observed that the attackers primarily converted the stolen assets into ETH.
③ Further money laundering risks: The FBI expects the attackers may further disperse the funds through DeFi and instant-swap platforms.
④ Interagency cooperation: The FBI, USSS, and IRS-CI urged exchanges, node operators, on-chain analytics firms, and other service providers to block transactions involving the relevant addresses.
The FBI also released the associated wallet addresses to help the industry track suspicious funds.
TRM Labs had previously attributed the incident to TraderTraitor; now, U.S. federal law enforcement agencies have officially announced their investigative findings.
The focus in the aftermath will shift from the source of the attack to asset tracking, preventing money laundering, and recovering the funds.
#Binance #Bitcoin #Crypto #Cybersecurity
FBI officially attributes the theft of over $6 billion in assets from Binance to a North Korean hacking group!
On October 31, the U.S. Federal Bureau of Investigation (FBI) issued a notice stating that North Korean malicious cyber activity known as TraderTraitor was involved in the October 29 theft from Binance.
The announcement highlighted four major points:
① Official attribution: The FBI attributed the attack to North Korea’s TraderTraitor hacking activity.
② Fund transfers: Investigators observed that the attackers primarily converted the stolen assets into ETH.
③ Further money laundering risks: The FBI expects the attackers may further disperse the funds through DeFi and instant-swap platforms.
④ Interagency cooperation: The FBI, USSS, and IRS-CI urged exchanges, node operators, on-chain analytics firms, and other service providers to block transactions involving the relevant addresses.
The FBI also released the associated wallet addresses to help the industry track suspicious funds.
TRM Labs had previously attributed the incident to TraderTraitor; now, U.S. federal law enforcement agencies have officially announced their investigative findings.
The focus in the aftermath will shift from the source of the attack to asset tracking, preventing money laundering, and recovering the funds.
#Binance #Bitcoin #Crypto #Cybersecurity
On October 31, the U.S. Federal Bureau of Investigation (FBI) issued a notice stating that North Korean malicious cyber activity known as TraderTraitor was involved in the October 29 theft from Binance.
The announcement highlighted four major points:
① Official attribution: The FBI attributed the attack to North Korea’s TraderTraitor hacking activity.
② Fund transfers: Investigators observed that the attackers primarily converted the stolen assets into ETH.
③ Further money laundering risks: The FBI expects the attackers may further disperse the funds through DeFi and instant-swap platforms.
④ Interagency cooperation: The FBI, USSS, and IRS-CI urged exchanges, node operators, on-chain analytics firms, and other service providers to block transactions involving the relevant addresses.
The FBI also released the associated wallet addresses to help the industry track suspicious funds.
TRM Labs had previously attributed the incident to TraderTraitor; now, U.S. federal law enforcement agencies have officially announced their investigative findings.
The focus in the aftermath will shift from the source of the attack to asset tracking, preventing money laundering, and recovering the funds.
#Binance #Bitcoin #Crypto #Cybersecurity
377
Log in to explore more content
Join global crypto users on Binance Square
⚡️ Get latest and useful information about crypto.
💬 Trusted by the world’s largest crypto exchange.
👍 Discover real insights from verified creators.
Email / Phone number
Trending Topics
NvidiaInvestigatesAfterHuang’sAffairApology
195,134 views
1,664 Discussing
Nvidia Board Opens Probe Into Whether Huang Was Target of Foreign Intelligence Effort
Binance News
·21 Likes
·54.9k views
BitcoinFiveWeeklyLosses
222,582 views
2,128 Discussing
AIChipSlumpNasdaq
120,880 views
1,313 Discussing