White House Eyes Pre-Release AI Vetting; Palantir Posts 85% Growth With U.S. Revenue Up 104%

The Trump administration is considering its first major AI governance reversal as Palantir reports the fastest growth in its history.

🏛️ White House Considers Mandatory Pre-Release Review of AI Models, With Anthropic's Mythos as Catalyst

Decoded: The Trump administration is evaluating a system that would give the federal government advance access to frontier AI models before they are released to the public, the New York Times reported May 4, confirmed by Reuters. The proposal stops short of blocking releases — it would give officials a review window, not veto power. The policy shift is being driven by concerns over Anthropic's Mythos model, which the company has withheld from public release after internal and external cybersecurity evaluations found it could materially accelerate complex cyberattacks, including identifying and exploiting vulnerabilities in government software systems. The NSA has already deployed Mythos internally to assess vulnerabilities in government Microsoft deployments. Anthropic previously described Mythos as too powerful for general public release — the first time a major U.S. AI lab has withheld a flagship model on those grounds. The proposed pre-release review process would represent a sharp reversal for the Trump administration, which has explicitly prioritized a deregulatory approach to AI and signed an executive order in January 2025 rescinding Biden-era AI safety requirements. (New York Times, May 4, 2026; Reuters, May 4, 2026)

Why it matters: A mandatory government review window for frontier AI models is the first serious governance mechanism the current administration has considered for the AI sector — triggered not by legislative pressure but by a specific capability threshold that cybersecurity officials concluded required federal access before public release. If enacted, the policy would give agencies like the NSA, CISA, and DoD first access to every new frontier model, creating a de facto national security clearance layer for AI releases. For AI labs, the review obligation would add time and disclosure requirements to model launches — and potentially expose model capabilities to government agencies before public benchmarks establish them. For investors, the shift signals that AI safety has crossed from a compliance topic into a national security priority the executive branch is now treating with operational urgency. The Mythos precedent — a model withheld not by regulation but by the developer's own risk assessment — is likely to accelerate this policy direction regardless of which administration implements it.

📊 Palantir Q1 2026: 85% Revenue Growth, U.S. Revenue Up 104%, Full-Year Guidance Raised to 71%

Decoded: Palantir Technologies reported Q1 2026 revenue of $1.633 billion — up 85% year-over-year and the highest quarterly growth rate in company history. U.S. revenue was the standout, growing 104% year-over-year as the company's AIP (Artificial Intelligence Platform) continued accelerating enterprise and government adoption. U.S. commercial revenue guidance was raised to 120% year-over-year growth for the full year. Palantir raised its full-year 2026 revenue guidance to $7.65–$7.66 billion — 71% year-over-year growth, 10 percentage points above the prior guidance issued last quarter and $380 million above the analyst consensus of $7.27 billion. Q2 revenue guidance was set at $1.8 billion, above the $1.68 billion consensus. Adjusted operating margin reached 60%; the company's Rule of 40 score — combining revenue growth rate and operating margin — hit 88%. CEO Alex Karp said the company doubled its U.S. business year-over-year. (BusinessWire, May 3, 2026; CNBC, May 4, 2026)

Why it matters: Palantir's 85% revenue growth at $1.633 billion is not the growth rate of an early-stage startup — it is an established public company with a decade of government contracts and enterprise software history delivering its fastest expansion ever, driven entirely by AI platform adoption. A Rule of 40 score of 88 is exceptional by any software benchmark: the metric combines growth and profitability, and a score above 40 is considered strong; Palantir more than doubled that bar. The 104% U.S. revenue growth confirms that the AIP platform — which deploys large language models inside enterprise and defense workflows — is converting to commercial revenue at an accelerating rate. For investors, Palantir's raised full-year guidance of 71% growth at $7.65 billion annualized makes it a real-time benchmark for how fast enterprise AI can scale inside a company that has both the cleared government infrastructure and the commercial AIP platform to serve both markets. The Pentagon AI classified contracts disclosed May 1 — which include OpenAI, Google, xAI, and Anthropic — introduce direct competition in the government segment Palantir has treated as a structural moat.

Stay decoded. See you tomorrow.

— The Get AI Decoded Team