MAGA hates AI, but Trump agrees with Bernie it might be time for partial government ownership

Trump said the U.S. government may take direct equity stakes in AI leaders OpenAI, Anthropic, and xAI, signaling a possible shift toward quasi-nationalization of frontier AI. The article highlights large projected losses at OpenAI ($44 billion through 2029, including $14 billion in 2026) and xAI ($6.4 billion operating loss on $3.2 billion revenue in 2025), with Anthropic the only company near profitability. The policy debate could materially affect AI valuations, capital structure, and regulatory expectations across the sector.

Analysis

The market implication is not “AI regulation,” but a creeping re-rating of frontier AI from a pure private venture to a quasi-utility with embedded political claims on future surplus. That shifts the distribution of outcomes: the upside to scale remains, but the probability of rent extraction, mandated sharing, or implicit public backstops rises sharply over the next 6-18 months, which is negative for private shareholders but positive for the firms that can convert political access into preferential financing, contracts, or regulatory insulation.

The second-order winner is not necessarily the large model lab itself, but adjacent platforms with durable enterprise moats and lower political toxicity. Names like IBM and Palantir are better positioned if the industry moves toward government procurement, compliance-heavy deployments, and “trusted AI” standards, while speculative frontier labs face a rising cost of capital and a greater chance of dilution or structured capital under government terms. On the loser side, the more cash-burning the model race gets, the more any public ownership discussion functions like a bailout signal, which can cap valuation multiples even before formal policy changes.

For hardware and critical inputs, the policy mix is subtly bullish for domestic supply-chain control but not for pure-play sentiment. MP and INTC could benefit if public policy shifts toward strategic national champions, yet the path is through slower, more bureaucratic capital allocation rather than clean free-market demand; that creates upside in the event of subsidies and procurement, but also execution risk and lower ROIC. The overhang is that any move toward equity stakes in AI may spill into compute, semis, and data-center infrastructure, broadening the set of “nationalization candidates” and compressing sector-wide multiples.

The key catalyst window is the next few months, not years: congressional signaling, executive-branch guidance, and any surprise funding/partnership language around frontier labs. The contrarian view is that the immediate price impact may be underdone on the winners and overdone on the most obvious targets—markets may be too focused on headline AI names and not enough on the beneficiaries of a regulated, government-backed procurement regime.

Key Decisions for Investors

• Long PLTR / short unprofitable frontier-AI proxies basket for 3-6 months: thesis is that government-linked AI demand and compliance spend benefit trusted platforms while private-lab dilution risk rises. Use a 1:1 notional pair; stop if Washington explicitly rules out ownership stakes.
• Add to IBM on 6-12 month horizon: asymmetric beneficiary if AI policy shifts toward regulated enterprise deployment and public-sector contracts. Target 15-20% upside versus limited policy downside; hedge with a sector short if needed.
• Tactical long INTC for 3-9 months only if policy rhetoric hardens into domestic industrial strategy. The setup is event-driven and headline-sensitive; size modestly because execution risk remains dominant.
• Small long MP as a secondary national-security beneficiary, but only as a basket hedge against AI-policy headlines. Better risk/reward in a broader strategic-minerals basket than in single-name size.
• Avoid chasing the most cash-burning frontier AI names until there is clarity on financing terms. If already long, consider buying downside protection into the next 1-2 policy catalysts, as any government stake language raises the odds of forced capital structure concessions.