How David Sacks crashed and burned in the White House

The Trump administration has shifted toward federal oversight of AI, including pre-deployment testing of commercial frontier models by CAISI, after earlier deregulatory efforts. The move was driven by national security concerns around powerful models like Anthropic’s Mythos, emerging foreign AI regulation, and geopolitical threats to critical data-center infrastructure, including reported Iranian strikes on AWS facilities in the UAE and Bahrain. David Sacks’ influence appears diminished as agencies regain authority and White House allies like Bessent, Lutnick, and Wiles take a larger role in AI policy.

Analysis

The market implication is not “more regulation” in the abstract; it’s a shift from ideology-led permissiveness to capability-led gating. That favors incumbents with the deepest compliance, testing, and government-relations benches because they can absorb pre-release scrutiny without derailing launch cadence, while smaller frontier labs face longer feedback loops and higher fixed costs per model iteration. The second-order effect is a widening moat for firms that can frame themselves as national-security partners rather than pure software vendors, especially where model evaluation, deployment controls, and audit trails become recurring service lines.

Security and infrastructure risk is the cleaner monetizable channel than the policy headline. If frontier models are now treated like dual-use systems, demand should accelerate for cyber-defense, model monitoring, identity/access controls, and regulated-cloud stack providers. At the same time, the geopolitics angle raises a non-obvious constraint: AI capex can be stranded by regional infrastructure fragility, so hyperscaler expansion and enterprise AI adoption in exposed corridors may face higher insurance, redundancy, and latency costs over the next 6-18 months.

The overdone part of the move is likely the assumption that this is uniformly bearish for Big Tech. For the largest platforms, regulation can actually reduce competitive churn and lower the odds of a destructive race-to-the-bottom on release quality; the winners are the firms already closest to the government testing regime and those selling the picks-and-shovels around it. The underdone risk is political volatility: a single change in White House personnel or a successful industry lobbying push could unwind the headline quickly, but the broader re-rating toward AI safety, cyber resilience, and data-center hardening should persist for quarters, not days.

Key Decisions for Investors

• Long MSFT / long GOOGL vs short basket of smaller frontier AI names not in the index: use as a relative-value expression that regulatory friction favors firms with the best compliance infrastructure; 3-6 month horizon, limited downside from revenue base, upside if CAISI testing becomes de facto standard.
• Initiate long PANW or CRWD as a policy-through-security hedge against frontier AI gating; 1-3 month entry on any AI selloff, with 15-20% upside if government-style model governance expands into enterprise budgets.
• Long AMZN puts or short AMZN on rallies for 1-2 quarter horizon: more exposed to geopolitical data-center disruption and infrastructure fragility, while valuation is vulnerable if Middle East/cloud resilience costs stay elevated; use defined-risk puts to cap carry.
• Buy NVDA downside protection via collars rather than outright short: policy pressure may slow some near-term deployment decisions, but demand is not broken; the better trade is to protect against multiple compression over 2-4 months rather than bet on a collapse.
• Pair long ORCL / short AMZN if CAISI-style approval frameworks shift spend toward regulated enterprise clouds and away from the most exposed hyperscale footprint; 6-month thesis with moderate convexity if government-adjacent workloads reprice in favor of compliance-heavy vendors.