finance
Trump Administration Considers Pre-Release AI Testing Rules
The Trump administration is reviewing a Biden-era proposal that would require pre-release testing and approval for certain AI models. The review follows a letter from more than 60 allies urging tighter oversight.
May 27, 5:30 AM·1m read1 source
FortuneAudio version
Tap play to generate a narrated version.
Developing·Limited corroboration so far. This page will refresh as more sources emerge.
The Trump administration is examining a 2023 regulatory framework that would require pre-release testing and approval for some AI models, according to reports cited by Fortune. The review comes after more than 60 of President Trump's allies sent a letter this week calling for greater government involvement in AI oversight.
On Thursday, President Trump postponed signing an executive order that would have increased oversight of the technology.
The approach under review centers on how AI systems are built and how they perform on tests before they reach the market. Fortune noted that similar rules appear in the European Union AI Act and in several state-level proposals in the United States.
These measures require developers to document accuracy, robustness, and data governance before a system is classified as high-risk and allowed to enter the market.
Fortune argued that capability-based rules cannot fully anticipate how AI systems will behave once deployed. The commentary proposed an AI Safety Management System that would emphasize continuous, real-world evaluation instead of point-in-time certification.
The piece outlined three principles: ongoing independent assessment of actual behavior, rules that target demonstrable harms while allowing innovation, and obligations scaled to a system's impact and autonomy. Fortune stated that the opinions expressed are solely those of the author and do not necessarily reflect the views of Fortune.
