We believe in a balanced, scientific approach where safety measures are integrated into the development process from the outset. This ensures that our AI systems are both innovative and reliable, and can deliver benefits to society.

At today’s AI Seoul Summit, we're joining industry leaders, government officials, and members of civil society to discuss AI safety. While there is still more work to do, we are encouraged by the additional Frontier AI Safety Commitments that OpenAI and other companies agreed upon today. The Commitments call on companies to safely develop and deploy their frontier AI models while sharing information about their risk mitigation measures, aligning with steps we have already taken. These include a pledge to publish safety frameworks like the Preparedness Framework⁠(opens in a new window) we developed and adopted last year.

This approach has enabled us to build and deploy safe and capable models at the current level of capability.

As we move towards our next frontier model, we recognize we will need to evolve our practices, in particular to increase our security posture to ultimately be resilient to sophisticated state actor attacks and to ensure that we introduce additional time for safety testing before major launches. We and the field have a hard problem to solve in order to safely and beneficially deliver increasingly capable AI. We plan to share more on these evolving practices in the coming weeks.