Big tech commits to voluntary frontier AI safety standards

At the end of September, President Trump and six Big Tech executives signed the White House Accord on Super Intelligence: Joint Commitment on Frontier Responsibilities. The document is not legally binding, though President Trump explained it is “morally binding.”

In just over 300 words, this voluntary pact places the responsibility of safely developing frontier AI models on the companies themselves. The pact states, “In order to build a positive future for the American people and the world, we believe every company is responsible for developing its own technology safely and in a way that builds trust with customers and the public.”

The pact emphasizes that each company training and deploying frontier models should implement “robust internal processes and controls to ensure that their technology behaves as intended and that any issues are promptly identified and resolved.” It then outlines four layers of controls and audits each company should establish.

1. Each company should establish “internal controls to monitor the capabilities and alignment of its models during training and deployment around areas like cybersecurity, biosecurity, and chemical threats, and to ensure that its models do not hack or access technical systems in unintended ways.”
2. Each company should have internal teams that are empowered “to ensure all of the controls, monitoring, and detection are operating as intended, and that any issues are remediated.”
3. Each company should work with “an independent external auditor or evaluator to carry out independent assessments of whether the controls, monitoring, and detection are operating as intended.”
4. Each company should “designate an independent committee of the board of directors to oversee and receive reports from the teams operating the controls and the internal and external auditors and evaluators, as well as to ensure any issues identified are remediated.”

The companies pledge to meet regularly “to establish standards and best practices to improve the safety of their systems.” These actions are voluntary for now, but the pact acknowledges that these actions may eventually be codified into law or regulation.

For those companies developing frontier AI models, it will be important to implement security controls and audits. While the above controls and audits are not mandatory, they could form the basis of future regulation. Getting these in place now can help companies stay ahead of any regulations that may be promulgated. Companies can assess their current layers of controls and audits to determine if any modifications or additions should be implemented.