The Trump administration is expected to release a new executive order focused on cybersecurity and artificial intelligence safety as early as this week, according to an Axios report citing sources familiar with the matter. The proposed order is aimed at strengthening protections around advanced AI models and creating a voluntary system under which AI companies would inform the government before releasing powerful new systems. The move comes as increasingly advanced AI models, including cyber-focused tools from companies such as Anthropic and OpenAI, have raised concerns inside the US government over potential cybersecurity risks.White House plan focuses on AI models and cybersecurityAccording to Axios, the executive order in its current form has two main sections — cybersecurity and what officials are calling “covered frontier models.”The cybersecurity section is expected to focus on improving security across national infrastructure, including hospitals, banks and government agencies. The report said the order would also aim to strengthen cybersecurity systems at the Pentagon and other national security offices, increase cyber hiring and encourage more information-sharing between AI companies and the government about security breaches.The second section would focus on advanced AI systems considered highly capable or risky. According to Axios, the government plans to create a review process to decide which AI models qualify as “covered frontier models” and assess them before they are publicly released.One of the key proposals in the draft order is a “voluntary framework” under which AI companies would share details about upcoming models with the government at least 90 days before public launch. The report said AI labs may also need to provide access to certain critical infrastructure providers.Axios reported that it is still unclear exactly which government agencies would oversee the framework, although both civilian and national security agencies are expected to play a role.The report said the Trump administration’s approach to AI safety has evolved in recent months after powerful new AI systems raised alarm among officials. AI models such as Anthropic’s Mythos and OpenAI’s GPT-5.5-Cyber reportedly demonstrated the ability to identify and exploit software vulnerabilities at very high speed, increasing concerns about cyber threats.The report further noted that cybersecurity was not initially a major priority for the administration. Earlier, the Trump administration reduced funding and staffing at the Cybersecurity and Infrastructure Security Agency.A White House official told the publication: “Any policy announcement will come directly from the president. Discussion about potential executive orders is speculation.”