OpenAI

OpenAI told two House Democrats that its engineers are developing “automated shutdown capabilities” for AI systems.

Guided by our National Security Principles, we approach national security deployments with safeguards, meaningful human judgment, and appropriate oversight, especially in high-stakes settings.

If safety is the priority, such a national standard is the best policy approach to ensure the US continues to lead in innovation and not be undermined by a patchwork of regulations.

First, the federal government should lead the testing and evaluation of the most advanced systems. Frontier AI raises national security and public safety questions that require technical expertise, resources, and access that no state can fully replic...

Nations should work together to develop compatible safety frameworks, trusted channels for sharing risk findings, and coordinated responses to incidents.

We also applied the method to challenging agentic rollouts, showing that it can extend beyond standard chat to more complex agent settings involving tool use, and can also be used for risk assessment before internal model deployments.

Transformative technologies can concentrate power, or they can broaden it. They can make life easier for a few, or they can expand opportunity for many. Our approach is rooted in the belief that AI should work for people: helping them pursue their ow...

One goal of such an organization should be to make it possible for the world to take coordinated action, including slowing frontier development when needed, so societal resilience, safety, and alignment can keep pace.

OpenAI has long believed there should ultimately be an international organization that helps coordinate leading AI efforts to reduce catastrophic risk. One goal of such an organization should be to make it possible for the world to take coordinated a...

We support strong, enforceable, risk-based regulations for teen safety and frameworks that pair robust safeguards with transparency and tools that put families in control. That includes requirements for privacy-preserving age assurance so companies c...

Policymakers should require the most capable frontier models to undergo a CAISI evaluation before public release. These evaluations should assess frontier capabilities, the effectiveness of associated safeguards and mitigations, and the resulting ris...

Establish a mechanism for companies to share information about incidents, misuse, and near-misses with a designated public authority. The system should emphasize learning and prevention over punishment, with appropriately scoped public disclosures th...

Today, we find that models’ reasoning is generally interpretable and easy to monitor. However, in the future, monitorability may break down for a variety of reasons.

This policy protects employees’ rights to raise issues, including about AI safety, applicable laws, or company policies.

In the US, we support state efforts to align around common frameworks such as California SB 53, the New York RAISE Act, and Illinois SB 315, which emphasize transparency, public reporting around catastrophic-risk evaluations and safety incidents, whi...

By default, we provide information and access to models that are intended to be public or production-ready. When the evaluations necessitate it, we provide deeper access, such as to helpful-only models or non-public information. OpenAI has provided t...

No single person or institution should define how an ideal AI should behave for everyone. To fulfill our mission of ensuring that AGI benefits all of humanity, OpenAI needs to build systems that reflect the wide range of values and priorities of all ...

We are likely to eventually need something like an IAEA for superintelligence efforts; any effort above a certain capability (or resources like compute) threshold will need to be subject to an international authority that can inspect systems, require...

First, we need some degree of coordination among the leading development efforts to ensure that the development of superintelligence occurs in a manner that allows us to both maintain safety and help smooth integration of these systems with society. ...

First, we need some degree of coordination among the leading development efforts to ensure that the development of superintelligence occurs in a manner that allows us to both maintain safety and help smooth integration of these systems with society. ...

First, we need some degree of coordination among the leading development efforts to ensure that the development of superintelligence occurs in a manner that allows us to both maintain safety and help smooth integration of these systems with society. ...

First, we need some degree of coordination among the leading development efforts to ensure that the development of superintelligence occurs in a manner that allows us to both maintain safety and help smooth integration of these systems with society. ...