Google, OpenAI, Anthropic in talks on AI safety
OpenAI has confirmed that it has been talking with rivals Anthropic and Google DeepMind about AI safety for several weeks. OpenAI's global policy chief Chris Lehane said this to reporters on Tuesday, according to TechCrunch.
According to insider reports from The Information, these talks have actually been going on since July. They reportedly focus on building a joint organization to oversee safety standards for AI systems.
The public debate was sparked by an essay from Anthropic CEO Dario Amodei published on Saturday, in which he called for the industry to slow down AI development to better catch up on safety. Days earlier, according to heise, an Anthropic employee had quit the company and publicly criticized it for not taking safety seriously enough.
Several AI leaders, including OpenAI's Sam Altman and Google's Demis Hassabis, backed Amodei's call. Altman said, according to TechCrunch, that OpenAI would bring in independent evaluators to monitor its safety processes.
It remains unclear, according to the reports, whether such cooperation between rival companies could violate antitrust law. Amodei proposed a government waiver for this, but Lehane said such a waiver was not necessary.
The Trump administration, according to TechCrunch, dismissed the safety concerns and warned that slowing down could give China an advantage. According to heise, Meta is also reportedly skeptical of a national oversight body.
It remains unclear whether and when the planned safety organization will actually be created, since much of the reporting relies on anonymous sources and the companies involved have only partially commented publicly.
WHY IT MATTERS
It shows how the biggest AI companies are handling the risks of their own technology and whether independent oversight might emerge.
WHAT YOU CAN DO NOW
- Stay critical of AI safety news Compare reports about AI safety across several reliable sources. This helps you tell whether a story is truly confirmed or based only on rumors.
- Check for independent oversight Look for whether an AI provider uses independent evaluators or auditing organizations. This tells you how seriously a company actually takes safety.
- Follow real regulation, not just announcements Check with official agencies or reputable media which AI laws or standards are actually being passed. Don't rely only on company statements.
- Use AI tools thoughtfully and sparingly Choose providers that are open about their safety measures. Think about what data you share and how the company handles it before you use a tool.