California report warns of unregulated AI’s ‘irreversible harms’

Without proper safeguards, powerful artificial intelligence (AI) could induce severe and ‘potentially irreversible harms,’ a report released Tuesday by California Governor Gavin Newsom warned.

The 53-page AI policy report from leading academics came as U.S. President Donald Trump advanced legislation that would impose a 10-year moratorium on state laws protecting against AI misuse.

The proposed federal bill would eliminate California’s existing bans on AI-generated child sexual abuse material, deepfake pornography, and robocall scams targeting elderly Americans.

‘This moratorium would mean that even if a company deliberately designs an algorithm that causes foreseeable harm, the company making that bad tech would be unaccountable to lawmakers and the public,’ warned the California report, citing opposition letters from more than 140 organizations.

Over 40 state attorneys general have joined the opposition, writing to Congressional leadership to express their concerns about leaving Americans vulnerable to AI-related threats, according to the California document.

Recent AI safety concerns have intensified as leading companies report their models approaching dangerous capability thresholds.

OpenAI said in February that risk levels were ‘medium’ across chemical, biological, radiological, and nuclear weapons categories, according to the California report.

Anthropic noted there is ‘substantial probability that our next model may require ASL-3 safeguards’ for systems that could help individuals create weapons of mass destruction.

Source: Xinhua

- Agencies