AI labs should pass safety review to get US government contracts, group says - Reuters

An advocacy group sent a letter to the Trump administration on May 11, 2026, recommending that AI labs pass a security review before their models are publicly released, and that failing labs be barred from receiving US government contracts. The letter was addressed to multiple officials within the administration, according to a report by Reuters. The recommendation comes amid ongoing debates about federal AI regulation.

The group argued that AI labs should undergo a security review covering topics such as model safety, cybersecurity, and responsible deployment. Labs that did not pass would be ineligible for government contracts. The letter did not specify which labs or contracts would be affected, nor did it detail how the review would be conducted. The advocacy group was not named in the report.

The Trump administration has generally favoured a light-touch approach to AI oversight, emphasizing innovation over restrictions. The advocacy group's letter represents one of several recent efforts by civil society to push for more safety requirements in government AI procurement. The proposal aims to ensure that AI systems used by the government meet minimum safety standards before they are deployed in public services.