White House Pivots to Stricter Artificial Intelligence Oversight With New Testing Agreements
Washington, Wednesday, 6 May 2026.
Driven by national security concerns over unreleased technologies, the US government has secured new agreements with major tech firms to rigorously test artificial intelligence models before public deployment.
A Strategic Reversal in Executive Policy
The current regulatory posture represents a stark departure from the Trump administration’s earlier technological directives. In 2025, President Donald Trump revoked a 2023 executive order established by the Biden administration that had required developers to share safety test results for high-risk AI systems [7]. Initially leaning into a deregulatory approach, the White House introduced an “AI Action Plan” designed to eliminate administrative burdens and foster rapid domestic innovation [6]. The Center for AI Standards and Innovation (CAISI)—originally launched as the AI Safety Institute under President Joe Biden—was subsequently restructured early in the Trump presidency to prioritize AI acceleration over safety constraints [1].