Bolt and Greater Paris plan autonomous driving pilot for 2028
The project will use Stellantis vehicles and assess transport service gaps, with commercial operations targeted for 2030.
OpenAI plans AI safety research with Poland’s NASK
The collaboration will examine how to assess more capable AI systems, find vulnerabilities and improve safeguards.
OpenAI said on October 23 that it would work with Poland’s NASK National Research Institute on technical research and the safety and security of artificial intelligence. The announcement comes amid growing scrutiny of unwanted behavior by AI agents.
Under the agreement, OpenAI and NASK’s AI Safety Research Center will investigate methods for assessing AI systems as their capabilities increase. Their work will also seek to uncover weaknesses and improve protective measures.
OpenAI’s approach to safety has drawn attention after it said a group of its AI agents broke out of containment and hacked the Hugging Face AI platform. Both OpenAI and Anthropic have disclosed examples of agents behaving in rogue or unintended ways.
OpenAI also said it has urged the United States to cooperate internationally on technical standards for frontier AI, meaning the most advanced systems. It called for that effort to draw on expertise from standards organizations and institutes focused on AI safety.