OpenAI says it is working with Anthropic and Google DeepMind on AI safety
According to a report by Bloomberg, the three companies have reportedly been in discussions for weeks on artificial intelligence safety measures, against a backdrop of growing concerns over the technology's risks.
OpenAI stated on September 15, 2026, according to a report by the Bloomberg news agency, that it is working with competitors Anthropic and Google DeepMind on issues related to artificial intelligence safety. The news currently comes from a single source; no independent confirmation is available.
OpenAI’s head of global policy, Chris Lehane, said that dialogue between the three companies has been ongoing for several weeks. The available sources do not specify exactly which safety issues are being addressed in the discussions, nor whether they have already produced shared documents or commitments among the three companies.
The stated collaboration involves three of the companies that develop some of the most widely used artificial intelligence models in the world, and comes at a time when the issue of economic and safety risks linked to the technology is receiving growing public attention, although the available sources do not specify what event or external pressure prompted the discussions among the three companies.
The fact that companies in direct commercial competition are stating that they are talking to each other about safety issues marks, in itself, a departure from the competitive posture that has characterized the sector so far, where labs tend to communicate progress and safety measures independently and separately. It is not clear, based on what has been reported so far, whether the discussions between OpenAI, Anthropic and Google DeepMind have a formal structure, a defined schedule, or a mandate that goes beyond informal exchanges between the respective policy officials.
It remains to be seen, in the coming weeks, whether the announced collaboration will translate into joint public initiatives, in line with what Lehane stated, or whether it will remain a private dialogue between the parties. Confirmation from sources independent of the three companies involved would allow for a more precise assessment of the actual scope of the initiative.
← Archive · Front page · Past editorials · Report an error · Original article (in Italian)