Tech News Roundup: Binance Hack, AI Security, Robotaxis, and Web Scraping Controls
The $6.3 Billion Binance Hack Was Pulled Off in Just 20 Minutes
Multiple crypto-tracing firms say the North Korea-linked TraderTraitor group may have carried out the record $6.3 billion Binance hack in about 20 minutes. Investigators believe much of the stolen cryptocurrency may never be recovered.
Read more: The $6.3 Billion Binance Hack Was Pulled Off in Just 20 Minutes
Binance Grapples With ‘Security Incident’ That May Cause $6.3 Billion in Losses - The New York Times
Binance is reportedly responding to a security incident that could result in $6.3 billion in losses. The available report does not specify the cause, affected assets, or whether the projected loss has been confirmed.
Read more: Binance Grapples With ‘Security Incident’ That May Cause $6.3 Billion in Losses - The New York Times
Silicon Valley wants you to invest in miracles
Anthropic’s 335-page S-1 filing presents AI development as potentially catastrophic for humanity or capable of ushering in an age of miracles. The stark framing illustrates how Silicon Valley is increasingly packaging uncertain technological futures to Wall Street investors.
Read more: Silicon Valley wants you to invest in miracles
Waymo’s next robotaxi is a purpose-built Toyota bZ4X
Waymo plans to launch robotaxi service in Tokyo in 2027 using Toyota bZ4X electric vehicles assembled in Japan. Waymo’s self-driving sensors and computers will be installed during production at Toyota’s Takaoka plant.
Read more: Waymo’s next robotaxi is a purpose-built Toyota bZ4X
A Supply Chain Attack Put Malicious Anthropic Claude Extensions on Thousands of Devices
A supply-chain attack placed malicious extensions in Anthropic Claude developer environments, reaching thousands of machines, but flawed code prevented the malware from spreading as intended. The incident highlights the risks of trusting third-party extensions.
Read more: A Supply Chain Attack Put Malicious Anthropic Claude Extensions on Thousands of Devices
The biggest names in AI are backing a new way to opt out of scraping
OpenAI, Google, Microsoft, and Anthropic say they will support a proposed standard that lets publishers exclude their sites from AI training across multiple services. The mechanism uses a policy token in robots.txt, building on the web’s existing crawler-control system.
Read more: The biggest names in AI are backing a new way to opt out of scraping