Trump AI Executive Order Draft Postponed Amid Concerns
A proposed executive order on artificial intelligence from President Donald Trump, which was anticipated for signing on Thursday afternoon, was unexpectedly put on hold. The document, a draft obtained by POLITICO, aimed to tackle worries that advanced AI products, such as those developed by companies like Anthropic, could be exploited to launch severe cyberattacks or cause other forms of disruption if they fell into the wrong hands. The order outlined a system for voluntary oversight, suggesting that developers of advanced AI models could submit their products to federal agencies for review up to 90 days before their public release, a detail previously reported by POLITICO.
The Proposed Oversight Framework
The core of the seven-page draft centered on establishing a voluntary review process for AI models. This framework would have allowed developers of sophisticated AI systems to present their creations to government bodies for examination well in advance of their market debut. The intention behind this early review was to preempt potential issues arising from the deployment of powerful AI tools. The draft specifically focused on advanced AI products, acknowledging their potential for both innovation and harm, particularly if misused by malicious actors.
Voluntary, Yet Questioned
A key aspect emphasized within the draft was the voluntary nature of these government AI reviews. The document explicitly stated that “Nothing in this section shall be construed to authorize the creation of a mandatory governmental licensing, preclearance, or permitting requirement for the development, publication, release, or distribution of new AI models, including frontier models.” This language aimed to reassure developers that the proposed system would not impose compulsory hurdles on their work. Despite the clear language, the idea that a voluntary system might one day become mandatory was, perhaps predictably, a point of concern. Former Trump AI czar David Sacks, for instance, voiced apprehension that the voluntary reviews could eventually transition into mandatory requirements, according to a senior White House official who spoke to POLITICO on Thursday.
Addressing Malicious Use
Beyond the oversight mechanism, the Trump AI executive order draft also contained provisions designed to combat the malicious use of artificial intelligence. It directed the attorney general to enforce existing federal statutes, specifically the Computer Fraud and Abuse Act, along with “other applicable Federal criminal laws.” This enforcement would target individuals who employ AI to illegally access or damage computer systems without authorization. Furthermore, it aimed to prosecute those who utilize AI while engaged in such unauthorized access to facilitate any other criminal activity, thereby addressing a range of potential digital offenses enabled by AI.
The Sudden Postponement
Leading up to the scheduled signing, administration officials had engaged with major technology companies, briefing them on the contents of the draft order earlier in the week. Top executives from the AI industry had also received invitations to attend a signing ceremony at the White House on Thursday. However, President Trump informed reporters that he had decided to halt the process. The president stated, “I didn’t like certain aspects of it,” and expressed worry that the order might impede the United States’ efforts to surpass China in the global competition to dominate artificial intelligence development.
The administration has not provided details on what specific changes might be made to the order, nor has it indicated when the signing might be rescheduled.