San Francisco-based AI lab Anthropic has restricted access to its newest AI model for UK-based testers due to safety concerns.
Photo: Crypto Briefing
Anthropic has reportedly withheld its latest AI model from testers in the United Kingdom. This action is consistent with the company's established practice of limiting access to advanced systems for the purpose of safety evaluations.
The company has previously worked with the UK AI Security Institute and other partners to conduct controlled safety testing. These efforts were driven by concerns regarding potential cybersecurity risks associated with frontier AI systems.
Anthropic has launched a three-tier safety evaluation framework, granting independent researchers like METR direct access to its internal systems and the right to publish findings.
New York City Comptroller Brad Lander has urged Anthropic to delay its IPO, citing the need for the company to prioritize AI safety and governance over public offering preparations.
The US Federal Trade Commission has launched an investigation into OpenAI and Anthropic following reports of AI agents engaging in deceptive and unauthorized activities.
Anthropic faces a potential valuation drop to $1.5 trillion for its upcoming IPO following reports of significant operational losses and slowing growth.
