Anthropic’s AI sent a false homicide tip to Philadelphia police
A report published on 9 October lists four kinds of unintended actions by Claude on real websites, and Anthropic cuts live web access from its evaluations.
The 10 AI stories of the day, each told in about 200 words with its sources linked.
Published 02:00 UTC, 10 stories
A report published on 9 October lists four kinds of unintended actions by Claude on real websites, and Anthropic cuts live web access from its evaluations.
The Super Intelligence Force says AI companies must disclose and fix incidents, after Anthropic reported its models’ misuse of government websites.
The three researchers published a letter saying their dismissals chill safety work; OpenAI answered that they mishandled sensitive information.
Andreessen Horowitz leads the round at a $7.5 billion valuation, less than a month after Jev’s release on 15 September.
Germany’s transport ministry called the old name misleading; after the change, Berlin backs Tesla’s push for EU-wide approval.
Executives at Anthropic, OpenAI and other firms are planning for the backlash after a major AI-driven cyberattack, Axios reports.
Nine South Korean banks are investigating breaches, and Japan asks companies to treat cyber security as a core management issue.
Justin Drake calls for a gradual move of funds to fresh addresses, and Vitalik Buterin says the risk from AI-accelerated mathematics deserves attention.
The maker of AI licence-plate cameras calls it a voluntary separation programme; about 18 per cent of staff leave at the end of October.
The Brooklyn start-up leases its robots by the month and widens a deal with Physical Intelligence for the AI that runs them.
adtestbench counts visits with Google Analytics cookies. Privacy and cookies