America's Cosmic Bet on AI
A Foreign Policy essay warns the US has become a high-stakes gamble on AI supremacy, with AI firms now driving most GDP growth, business investment, and 85% of 2026's S&P 500 gains.
Intelligence analysis by Llama
Harvard's Graham Allison argues the United States has wagered its economy, fiscal solvency, and national security on winning the AI race against China, raising the question of what happens if the bet fails.
Imagine if your family's entire savings were put on a single game at a casino. That's kind of what the United States is doing with AI — betting most of its economy, its stock market, and even its military power on the hope that AI companies will win the global race. If they win, great. If they lose, a lot of people could lose a lot of money.
Analysis
The Economy Has Become a One-Industry Wager
The most striking data point in Allison's essay is the degree to which AI has become the load-bearing pillar of US economic growth. According to the piece, AI-related investment is now driving most of US GDP growth, and spending on AI data centers accounts for half of all business investment in the country. AI companies made up more than 40 percent of US equity market value before SpaceX's June 2026 IPO, and if OpenAI and Anthropic follow with their expected trillion-dollar-plus IPOs, AI firms will account for more than half of US stock market value. Of the S&P 500's gains in 2026, 85 percent have come from AI companies, per the article.
This is no longer a sector story — it is a macro story. A single technology has become the primary engine of corporate investment, equity returns, and GDP expansion. The diversification that traditionally cushions an economy against sector-specific shocks is rapidly eroding. Allison's rhetorical question, asking whether a fiduciary investor would place this much capital in a single sector, applies not just to a portfolio manager but to the country itself.
A Tripartite Coalition and the National-Security Frame
The political economy around AI has produced an unlikely alignment of interests. Allison highlights an "odd alliance" including President Donald Trump, Senator Bernie Sanders, and OpenAI CEO Sam Altman that is seeking ways for the US government to become a major stakeholder in leading AI companies. The national security apparatus — the Defense Department, CIA, and NSA — is becoming a major customer. David Sacks, who recently stepped down as Trump's AI czar, warned in late July that OpenAI and Anthropic are pursuing a "regulatory capture strategy" aimed at handicapping competitors, both foreign and domestic.
Three forces are propelling this dynamic: the unprecedented financial upside for AI company leaders, the desperate fiscal pressure on the US Treasury (annual deficits near 6 percent of GDP, with interest payments now exceeding the defense budget), and a national security consensus that whoever builds artificial general intelligence first will have a decisive, irreversible advantage. As former Secretary of State Condoleezza Rice put it, this is a race the US "absolutely have to win." That framing turns AI leadership from a commercial question into an existential one — and removes the usual policy guardrails.
Why the House Could Lose
Allison's central question — what happens if the US loses? — is left partially open in the truncated text, but the warning is unmistakable. A strategy that ties GDP growth, equity market value, fiscal stability, and national security to a single emerging technology invites compounding failure modes: a bubble collapse that wipes out half the stock market, an AGI breakthrough by a rival that strands US investments, or a productivity miracle that fails to materialize and leaves the Treasury without a lifeline for its deficits. The article's final visible question — whether the US has chosen a prudent course — implicitly answers no.
The deeper implication is that the AI race has reorganized American capitalism around a single bet. Whether that bet pays off or not, the concentration itself has already changed the relationship between the US government, the tech industry, and the financial markets in ways that will outlast any individual policy or administration. The essay's title frames it as a "cosmic" wager, and the data it marshals suggest the stakes match the rhetoric.
Key points
- AI companies now drive the majority of US GDP growth, half of all business investment, and 85 percent of S&P 500 gains in 2026, per the article
- OpenAI and Anthropic are expected to pursue trillion-dollar-plus IPOs later this summer, which would push AI firms past half of US stock market value
- An unusual Trump-Sanders-Altman alliance is seeking US government equity stakes in leading AI companies, while the Pentagon, CIA, and NSA become major customers
- Former Trump AI czar David Sacks warned that OpenAI and Anthropic are pursuing a 'regulatory capture strategy' to handicap competitors
- US fiscal deficits near 6 percent of GDP and rising debt service costs make an AI productivity breakthrough attractive to policymakers seeking to avoid an economic catastrophe
An AI productivity miracle could help the US Treasury escape its fiscal trap, with rising growth and tax revenues offsetting deficits that are now near 6 percent of GDP. Winning the AGI race, as national security strategists frame it, could deliver decisive military and economic advantages that cement US global primacy for decades. A close public-private partnership with AI leaders could also ensure the technology is deployed in ways that align with democratic values and national interests.
An AI bubble burst could erase most US equity gains and trigger a recession given that AI firms make up more than 40 percent of the stock market and 85 percent of the S&P 500's 2026 gains. Losing the AGI race to China could leave the US with stranded investments, eroded military advantage, and a fiscal crisis without a productivity lifeline. Regulatory capture by incumbents, as David Sacks warned, could also stifle competition and entrench a small group of companies, concentrating both economic and political power.
