Anthropic warns AI poses catastrophic risks ahead of blockbuster US share listing

Warnings in the listing filing

In documents filed ahead of its planned listing, Anthropic said concerns had been voiced about AI systems posing significant risks to humanity. The company said those concerns came from policymakers, regulators and researchers, as well as people within its organisation.

Anthropic also repeated its commitment to developing AI in ways that benefit people, while warning investors about potential dangers. The company said it would aim to be “bold but sane” and build trust through successful decisions, according to its filing.

A closely watched market debut

Anthropic announced in June that it intended to go public. The company was founded five years ago by chief executive Dario Amodei, his sister Daniela Amodei and five other co-founders.

The number of shares to be released, their expected price range and the date of the listing have not been disclosed. Anthropic’s debut would make it the first true AI company to go public and could serve as a test case for the technology’s future.

AI investment and security concerns

The plans come amid heavy spending by technology companies on AI infrastructure and tools, including data centres and the models underlying AI technology. More than $1tn was invested in those efforts over the last year alone. After the companies had decided on such plans, a wave of cyber-security and other technological breaches occurred involving AI agents, or bots designed to operate somewhat autonomously.