OpenAI CEO Sam Altman pitches 3 paths to control AI
OpenAI CEO Sam Altman just pitched three possible paths to make sure AI doesn't get out of control.
He's considering: setting up an international group (kind of like the IAEA for nuclear safety) to enforce global AI rules, merging major AI firms into one team effort, or letting one developer gain enough economic power to cut off rivals' access to computing resources, which he's not a fan of.
Altman told his team that OpenAI must decide within the next year which of these paths it will back.
Labs, draft bill address AI risks
These ideas come as worries grow about security leaks in AI from companies like OpenAI, Google, Anthropic, and Meta.
To tackle this, several top labs are working on a new Standards Authority for Frontier AI, set to launch by early 2027.
There's also talk of a draft bill that could allow the Commerce Department to halt models capable of causing large-scale harm.