AI Giants Unite for Safety Standards
Leading artificial intelligence companies in the sector are preparing to establish a new standards organization to audit their own models without government oversight.
Anthropic, Google, and OpenAI have started working together to establish a new standards body that will provide self-regulation for artificial intelligence models.
Goals of the New Organization
Anthropic, Google, and OpenAI aim to establish a standards organization where leading companies operating in the field of artificial intelligence can self-regulate without government oversight. The new structure is planned to be launched at the end of the year or early next year.
Shift in Public-Private Initiative
The companies initially aimed to establish a public-private partnership, but these efforts reportedly stalled during the Trump administration. Consequently, work began on an independent structure where developers will determine their own safety commitments.
Planned Responsibilities of the Body
The planned standards organization will support third-party entities that test models prior to deployment. It will also detail how artificial intelligence developers should report safety incidents and define qualifications for independent auditors.
Past Similar Initiatives
Anthropic, Google, OpenAI, and Microsoft launched a non-profit organization called the Frontier Model Forum in 2023 with similar goals, and this group is known to continue its activities.
Criticisms and Competition Concerns
There are also criticisms regarding the new organization plan. Some circles worry that the three major artificial intelligence companies could use this structure to exclude open-source model developers and other competitors.