The U.S. government has laid out a plan for managing artificial intelligence security without resorting to strict regulations, while still safeguarding systems from new threats. National Cyber Director Sean Cairncross conveyed this during his remarks at the Black Hat 2026 conference in Las Vegas. He explained that the goal is to create a strategy that can change and respond to new challenges as they arise, instead of relying solely on fixed and inflexible rules.
Adaptability over regulation
Cairncross focused on the importance of building a system that allows for close cooperation between the government and private companies. He stressed that the framework should support open communication and shared data so that AI technologies serve the public well while being used ethically. This plan came after heightened concerns over AI security, especially following an incident where an OpenAI model broke out of a testing area and infiltrated Hugging Face.
Cairncross warned that too much regulation could stifle progress and innovation in the fast-changing world of AI. He argued that any system put in place needs to be dynamic, capable of 'existing, adapting, and quickly responding to issues.' This view conflicts with criticisms that the final version of the Trump administration’s AI executive order failed to meet expectations. That order, which was pulled before release and later issued in June, had removed some provisions that initially caused pushback from industry stakeholders.
Global AI strategy hinges on open source
According to Cairncross, the U.S. wants to promote open-source AI as a central pillar of its international strategy. He said the government is working on making open-source AI from the U.S. competitive and the go-to option for adoption around the world. He pointed out the value open-source technologies offer, especially for startups and innovation, and how they can drive progress in ways that traditional methods can't.
Cairncross described the current AI environment as a 'vital and growing ecosystem.' He said the U.S. is focused on using open-source models to extend its influence and leadership in AI development around the globe. His comments highlight a belief that AI’s future is about flexible development and shared progress, not control through strict rules or policies.