The Trump administration is poised to issue an executive order addressing cybersecurity and artificial intelligence (AI) safety, sources familiar with the matter confirmed. The order, expected as early as this week, seeks to establish a voluntary framework for AI developers to inform the government of new models prior to their release, ensuring national security interests are safeguarded.
Key Components
The draft order focuses on two primary areas: cybersecurity and 'covered frontier models.' The cybersecurity measures aim to bolster defenses for critical institutions such as hospitals, banks, and national security agencies, while also promoting cyber hiring and information sharing between the government and the AI industry. The frontier model component introduces a multi-layered review process to determine which AI models fall under the 'covered frontier' category, requiring pre-release assessments.
'Any policy announcement will come directly from the President. Discussion about potential executive orders is speculation,' a White House official stated.
Voluntary Framework
The draft proposes a voluntary framework under which AI labs would share their models with the government at least 90 days before public release. While specifics on enforcement agencies remain unclear, both national security and civilian agencies are expected to play roles in the process. This initiative reflects the administration’s attempt to balance AI innovation with growing concerns over cyber threats and AI safety.
The move comes as advanced AI models, like Anthropic's Mythos, have raised alarms within the government, prompting a reassessment of the administration’s previously hands-off approach to AI regulation.