Hassett rules out new agency for AI oversight By Investing.com

The Trump administration is not planning to create a new agency to oversee advanced AI models, according to White House NEC Director Kevin Hassett. Instead, officials are coordinating with the private sector to reduce the risk of significant harm before public release, while considering whether to issue an executive order that could expand existing cybersecurity information-sharing programs to include AI companies. The update is policy-oriented and incremental, with limited immediate market impact.

Analysis

This is a meaningful negative surprise for the most regulation-sensitive corner of AI infrastructure: the market had begun to price a higher-probability federal gating mechanism that would favor incumbents with compliance budgets and punish smaller frontier model developers. Pulling back from a new agency reduces the odds of a near-term permit/tax regime, which is mildly bearish for the “AI regulatory scarcity” premium embedded in some large-cap platform names, but supportive for the broad AI tape because it delays any forced re-rating of model training economics.

The more interesting second-order effect is that the policy burden may shift from pre-release model governance to cybersecurity and information-sharing, which is structurally better for security vendors than for model labs. If the administration leans on existing cyber channels, budget dollars should migrate toward monitoring, identity, data-loss prevention, and incident-response layers rather than toward dedicated AI compliance tooling; that favors the picks-and-shovels stack over pure-play model monetization. It also lowers the probability of a uniform federal standard, which means state-level and sector-specific rules remain the main drag, creating a fragmented compliance environment that larger incumbents can absorb more easily than venture-backed challengers.

The contrarian read is that “no new agency” is not the same as “no regulation” — it may actually be more bullish for the biggest incumbents because informal coordination and selective enforcement are harder for smaller firms to navigate. The key catalyst to watch over the next 1-3 months is whether the White House codifies this through an executive order; absence of formal action should keep policy volatility low, but any draft centered on cybersecurity information-sharing would be a near-term catalyst for cyber names and a modest headwind for AI governance software.

From a risk perspective, the biggest reversal would be a headline-triggered AI safety incident, which could compress timelines and force rushed regulation within days. Absent that, the policy overhang likely fades over weeks, and the market will refocus on capex and product cycles rather than rules, which is constructive for the mega-cap AI complex but less so for companies selling compliance narratives.

Key Decisions for Investors

• Add to long MSFT/NVDA on any intraday weakness over the next 1-2 sessions; the absence of a new agency reduces near-term regulatory convexity, and these names still have the balance-sheet scale to absorb future compliance costs.
• Initiate a pair trade: long PANW or CRWD vs. short a basket of high-beta AI governance/compliance software names for 1-3 months; if policy flows through cyber channels, the spend should accrue to established security platforms rather than niche AI oversight vendors.
• Sell 1-2 month out-of-the-money puts on QQQ or SMH only if implied vol remains elevated; the setup favors volatility bleed as regulatory fear decays, with limited downside unless an AI safety incident becomes a fresh catalyst.
• If an executive order draft is released, buy short-dated call spreads in PANW/CRWD into the announcement and trim AI model exposure; the first-order beneficiary would be cyber infrastructure, not model developers.
• Avoid initiating fresh shorts in frontier AI names solely on this news; the policy outcome is more deferential than punitive, so the better risk/reward is to wait for an actual rulemaking or incident before pressing downside.