The White House is floating a plan to vet artificial intelligence the way the FDA vets drugs. National Economic Council Director Kevin Hassett confirmed Wednesday the Trump administration is weighing an executive order requiring a formal AI evaluation process before models reach market. Food-tech operators: this one is worth watching.
TLDR
- White House may require pre-market AI review, similar to FDA drug approval.
- NEC Director Kevin Hassett confirmed the executive order is under consideration.
- A new AI model called Mythos is reportedly driving the administration’s urgency.
- Food-tech and ag-tech suppliers using AI tools could face new compliance layers.
- No timeline confirmed; the policy is still in early formation stages.
An AI Evaluation Process Modeled on the FDA
Kevin Hassett, director of the National Economic Council, told reporters Wednesday the administration is actively considering the move. The trigger: a new advanced AI model called Mythos, which has accelerated internal debate about AI oversight. Hassett offered few specifics on scope or timeline. However, the FDA-drug analogy signals a preference for structured, evidence-based review before deployment.
For food-industry operators, the parallel is not abstract. AI tools already run demand forecasting, ingredient sourcing, and quality control across the supply chain. A formal AI evaluation process could require vendors to document model performance, bias testing, and safety thresholds before selling into regulated industries.
What Food-Tech Suppliers Should Track Now
No executive order has been signed. Specifically, no agency has been named to administer any review. But the FDA comparison matters: it implies a pre-market clearance model, not a post-incident enforcement model. That is a significant structural difference.
Suppliers building AI into food safety systems, nutrition labeling tools, or traceability platforms should monitor this closely. The Hill’s reporting notes the White House is still in early deliberation. Early. But the direction is clear: the federal government is moving toward treating high-stakes AI like a regulated product, not a software update. Operators who build compliance readiness now will not be scrambling later.
Source: The Hill. https://thehill.com/policy/technology/5866292-white-house-ai-evaluation-process/