Will the US Take an Equity Stake in Anthropic by 2027?

---

Anthropic's Prediction Market Odds Drop Even as US Government Deepens Its AI Bet

The US government has no existing legal framework or recent precedent for taking equity stakes in private AI laboratories. That single structural fact is reshaping how prediction markets price the possibility of Washington acquiring ownership in Anthropic, even as the San Francisco-based AI company cements itself as one of the federal government's most strategically important technology partners.

Over the past three days, Anthropic's implied probability in the "Which companies will the US take a stake in before 2027?" market has fallen from 42% to 33%, a 9-percentage-point decline that places the company at its period low. The drop is visible across both major platforms: Kalshi prices Anthropic at 28%, while Polymarket holds at 38%, a 10-point spread that signals disagreement about whether the structural barriers to a government stake are temporary or permanent.

The paradox is real. Anthropic has been accumulating federal AI contracts, positioning its Claude model family as a trusted tool for government applications, and building a Washington footprint that rivals any technology company a fraction of its age. Under conventional logic, deeper government entanglement raises the probability of an equity arrangement. Markets are saying the opposite, and the reasoning is worth interrogating.

---

What a US Government Equity Stake in Anthropic Would Actually Look Like

The United States government has taken equity positions in private companies before, but the historical pattern is narrow and specific. During the 2008–2009 financial crisis, the Troubled Asset Relief Program (TARP) authorized the Treasury to acquire preferred stock and warrants in banks, insurers, and automakers. The government held equity in General Motors from 2009 to 2013 and in AIG from 2008 to 2012. In both cases, the mechanism was a bailout: the companies were failing, and equity was the price of survival.

The closest modern analog outside of crisis intervention is the CHIPS and Science Act of 2022, which authorized direct investments in semiconductor fabrication facilities. But even those arrangements were structured as grants and loans with clawback provisions, not equity stakes. The recipients were manufacturers building physical plants on US soil, companies like Intel, TSMC, and Samsung with tangible assets the government could collateralize.

Anthropic fits neither template. It is not failing. It is not building fabrication plants. Its primary assets are trained neural network weights, research talent, and intellectual property. For the US government to take an equity stake in Anthropic before December 31, 2026, it would need to invoke an existing but untested authority, secure new legislative authorization, or structure a novel arrangement through a national security vehicle like the Defense Innovation Unit or In-Q-Tel. Each pathway requires months of legal groundwork that, as of July 2026, has not been publicly initiated.

---

Federal Contracts, National Security Framing, and Anthropic's Growing Washington Footprint

None of this means the relationship between Anthropic and the US government is cooling. The opposite is true. Anthropic has positioned itself as the "safety-first" frontier AI lab, a framing that resonates with policymakers concerned about both adversarial misuse and competitive pressure from Chinese AI development. The company's leadership has consistently engaged with Congress, the White House Office of Science and Technology Policy, and national security agencies in ways that go beyond typical corporate lobbying.

Anthropic's federal AI contract activity has expanded across defense and intelligence applications. The company's Responsible Scaling Policy, which commits to safety evaluations at defined capability thresholds, has given government partners a framework for risk management that competitors have been slower to formalize. This policy engagement creates the conditions for deeper partnership: classified deployments, cleared personnel, and eventually the kind of strategic dependency that could motivate an equity arrangement.

The bull case for Anthropic in this market rests on a simple premise: if the administration decides that frontier AI capability is critical national infrastructure, comparable to semiconductor fabrication or nuclear energy, then the precedent for government ownership exists in spirit even if the legal mechanics are undeveloped. A 33% implied probability means the market assigns roughly one-in-three odds to that logic prevailing before year-end.

---

The Case Against: Why Markets Are Right to Fade This

The strongest argument against a US equity stake in Anthropic before 2027 is not about Anthropic's relationship with government. It is about time and mechanism. Six months remain before the resolution deadline. No bill authorizing AI equity stakes has been introduced in either chamber of Congress. No executive order has established the legal basis for such an arrangement. The Committee on Foreign Investment in the United States (CFIUS) reviews inbound foreign investment; it does not authorize outbound government equity purchases.

Even in crisis scenarios where the government has moved quickly on equity, the timeline from authorization to execution was measured in months. TARP was signed into law on October 3, 2008, and the first bank equity purchases closed on October 28, a 25-day sprint enabled by emergency legislation and existential financial panic. No comparable urgency exists in the AI sector today. Anthropic is well-capitalized, having raised billions from Amazon and other investors.

The 10-point spread between Kalshi (28%) and Polymarket (38%) may reflect this disagreement in microcosm. Kalshi's lower price suggests its trader base is more skeptical of structural feasibility, while Polymarket's higher number may incorporate a broader interpretation of what constitutes a "stake," including indirect arrangements through sovereign wealth-adjacent vehicles.

---

Tracking the Decline: Anthropic's Equity Stake Odds Over Time

The three-day chart tells a story of steady erosion rather than a single catalyst-driven collapse. No public announcement, legislative development, or policy reversal appears to have triggered the move. Instead, the decline from 42% to 33% looks like the market gradually absorbing the mechanical reality: the resolution deadline is approaching, and the prerequisite legal and legislative steps have not materialized.

At 33%, the market is not dismissing the possibility. It is saying that the path from "deepening strategic relationship" to "actual equity ownership" requires a series of steps, each individually plausible but collectively improbable within six months. The government can award Anthropic every classified AI contract available without acquiring a single share of equity. Contracts and ownership are different instruments serving different purposes.

The market's current pricing is defensible. Anthropic's government ties are real and growing, but ties are not stakes. Until Congress authorizes a mechanism, or the executive branch identifies an existing one, the gap between partnership and ownership remains wide. Traders who bought at 42% and are watching 33% need to ask themselves a concrete question: which specific legal authority will the US government invoke, and when? If they cannot name one, the market is telling the right story.