Elon Musk Says AI Will Double US GDP Growth to 4% Next Year — Economists Are Not Convinced

Elon Musk wants you to believe the American economy is about to double its speed. On Friday, September 18, the Tesla and xAI chief posted a one-sentence forecast on X: "My guess is that AI roughly doubles US GDP growth next year from ~2% to ~4%. Maybe even more."

Read that again slowly. Doubling GDP growth in a single year would rank among the sharpest economic accelerations in modern American history — the difference between an economy cruising at its speed limit and one suddenly doing twice the limit with the same engine. It is the kind of claim that stops a newsroom cold. It is also the kind of claim that deserves to be checked against the people who forecast GDP for a living.

So let's do that. Here is what Musk said, what the official numbers say, where his argument has genuine force — and where it falls apart.

Elon Musk speaking at an event in National Harbor, Maryland, in February 2025. Photo: Gage Skidmore, via Wikimedia Commons (CC BY-SA 3.0).

The Friday post

There is not much to the post itself, and that is part of the point. One sentence, no data, no model, no chart. Just a guess — Musk's word, not ours — that artificial intelligence will lift American growth from roughly 2% to roughly 4% next year, "maybe even more."

It would be easy to dismiss a one-line social media post as noise. It would also be a mistake. Musk's economic pronouncements move markets and headlines with unusual force, partly because he runs companies at the center of the AI buildout — Tesla, xAI — and partly because his predictions have a way of becoming the week's talking points whether economists like it or not.

The timing sharpened the attention. The post landed on a Friday. The next day, Saturday, September 19, President Donald Trump took to Truth Social to announce he is forming an "AI Force," predicting that AI could eventually account for "possibly as much as 25% of our Country's GDP." There is no evidence the two posts were coordinated. But together they turned one weekend into an unofficial referendum on a single question: how big does AI get, and how fast?

The record Musk is arguing against

Start with the baseline, because the baseline is brutal to Musk's case.

According to the U.S. Bureau of Economic Analysis, real GDP grew at a 1.5% annual rate in the second quarter of 2026 — down from 2.1% in the first quarter. The quarter before that, the final three months of 2025, managed just 0.5%. That is three consecutive readings that describe an economy losing momentum, not gaining it.

Then there is the Federal Reserve. In projections released September 16 — two days before Musk's post — the central bank penciled in 2.0% growth for 2025 and a 2.0% long-run growth rate: its estimate of the economy's cruising speed once the cycle settles. In plain terms, the Fed believes the American economy grows at 2% a year when nothing unusual is happening.

Musk is predicting 4%. That is not a tweak. That is growth at double the central bank's estimate of the economy's speed limit, sustained for a full year, starting from an economy that just printed 1.5%. For that to happen, something would have to inject roughly two extra percentage points of growth into a roughly $30-trillion economy — the equivalent of conjuring an entire new sector out of thin air, in twelve months.

That "something," in Musk's telling, is AI. So the next question is whether the mechanism exists.

Where the growth could actually come from

This is where Musk's argument has its strongest leg to stand on — and it is worth taking seriously.

The AI boom is not an abstraction. It is concrete, steel, copper, and silicon: data centers rising across Virginia, Texas, and Arizona; orders for chips and power equipment stretching years into the future; construction crews and engineers hired by the tens of thousands. All of that spending shows up in GDP the moment the money is spent. You do not need AI to make anyone more productive for data-center construction to lift growth — you just need the checks to clear.

And the scale is genuinely historic. Apollo Global Management chief economist Torsten Slok noted on August 6 that hyperscaler data-center spending — Amazon, Microsoft, Alphabet, Meta, and Oracle — has climbed from 0.3% of GDP in 2019 to 1.4% in 2025. Consensus forecasts, Slok added, hold that share near 3% of GDP every year from 2027 through 2029. A single industry's capital spending approaching 3% of the entire American economy is without modern precedent.

Cooling towers and backup generators on a data center roof. Hyperscaler data-center spending has risen from 0.3% of US GDP in 2019 to 1.4% in 2025. Photo: Rsparks3, via Wikimedia Commons (CC0 public domain).

That is the transmission mechanism, and it is real. Every percentage point of GDP redirected into data centers is a direct addition to growth while the building lasts. Musk's 4% is not conjured from nothing; it is an extrapolation of a boom you can photograph.

But extrapolation is where the trouble starts. Construction spending is a level, not a growth rate that compounds forever — once the data center is built, you need to build another one to keep the contribution going. And 4% economywide growth needs more than one hot sector. It needs the productivity gains — actual AI-driven efficiency spreading across every industry — to arrive on schedule. That part has not shown up in the data yet.

The Dolat Capital test

Which brings us to the most clear-eyed piece of analysis published on this boom so far. In a recent report, the brokerage Dolat Capital argues the AI investment cycle is entering its "first meaningful macro test" — and lays out exactly what could break.

The core observation: the hyperscalers have quietly transformed from asset-light, cash-returning businesses into heavy industrial spenders, funding enormous capital programs with a mix of cash flow, debt, and equity. That works beautifully when money is cheap. It works less beautifully when central banks turn more hawkish and the funding hurdle rises — which is what Dolat sees happening, with bond yields climbing amid heavy government borrowing, Japan's policy normalization, and a global wave of infrastructure and defense spending competing for the same capital.

Then there is the revenue side of the ledger. Monetization, Dolat argues, remains the biggest unresolved issue in the AI trade: token costs are falling, models are getting more efficient, the technology is changing fast, and each successive model generation gets a shorter window to earn back its development cost. The money going in is enormous and certain. The money coming out is enormous and hypothetical.

Dolat's summary line deserves to be quoted in full: "The key risk is not demand for AI, but whether incremental investment continues to generate sufficient returns to sustain the current pace of spending."

That sentence is the hinge of this entire story. If the returns materialize — if enterprises actually pay for AI at a scale matching the buildout — then the capex boom sustains itself and Musk's growth story has a foundation. If they do not, the spending slows, the construction crews go home, and the growth forecast goes with them. Musk's post assumes the first outcome. Dolat is pointing out, correctly, that it is an assumption, not a fact.

The Christmas downgrade

There is one more piece of context Musk did not mention, and it is the most revealing of all.

On December 24, 2025, Musk posted a different forecast: "Double-digit growth is coming within 12 to 18 months." That window runs through June 2027 — it is still open. Double-digit means 10% or more.

His new forecast is 4%. That is less than half of what he promised nine months ago. As TheStreet noted, the new call trims the target by more than half — without saying so. Nobody held a press conference to announce the revision. It just showed up, one Friday, in a single sentence, as if the old number had never existed.

This is a pattern worth naming. Musk's forecasts tend to arrive with enormous confidence and depart with enormous quiet. The honest way to read the 4% call is not as a careful economic projection but as a vibe — a restated, downsized version of a bigger promise that is quietly expiring. It may still prove directionally right; AI investment really is transforming the growth picture. But a forecaster who moved the goalposts from 10% to 4% without comment has not earned the benefit of the doubt on the new number.

And there is an awkward wrinkle. Earlier this month, Musk publicly backed Anthropic chief Dario Amodei's call for the industry to slow frontier AI development — posting "Dario is right," according to AP and CNN reporting — even as his own company, SpaceXAI, was named among the four labs in the federal antitrust lawsuit accusing the top AI labs of agreeing to slow development. Slowing down and doubling growth are not strictly contradictory — you can pour concrete while pausing model training — but it is an odd posture: the man predicting an AI-driven economic miracle is also on record saying the industry building it should tap the brakes.

Why this matters to you

Strip away the personalities and this is a story about your paycheck, your prices, and your power bill.

On jobs, the two scenarios diverge sharply. If Musk is right and AI delivers a genuine productivity boom, history suggests the gains eventually show up in wages — the 1990s tech boom is the template. If Dolat is right and this is a capex cycle that outruns its revenues, the near-term story is construction jobs now and a potential overhang later. Either way, the labor market of the next two years will be shaped less by AI replacing workers than by AI's physical buildout employing them.

On inflation, the picture is genuinely uncertain. Growth at 4% with the economy's productive capacity rising to meet it is the dream scenario — more output without higher prices. Growth at 4% driven by deficit-funded spending and constrained power grids is the inflationary one. Note that prices, as measured by the GDP price index, were already rising at more than a 5% annual rate in the second quarter. A growth boom layered on top of that is not obviously benign.

Then there is the power bill. The data-center boom consumes electricity at a scale utilities were not built for, and households are starting to notice. On September 18, Reuters reported that Trump was discussing legislation with Senate Majority Leader John Thune to shield households from data-center-driven electricity costs — with Trump calling data centers "the oil of the next 20, 25 years." The growth Musk is forecasting has a meter attached to it, and voters will see the number every month.

Finally, the politics. With the midterm elections approaching, the state of the economy is always the central campaign issue, and both parties will now be tempted to claim the AI boom as their story. The analytical point — not a partisan one — is that growth forecasts have become political instruments. A 4% promise is useful to a White House selling an economic narrative whether or not it materializes. Voters should watch wages and prices, not headlines.

The honest verdict

Let's separate what we know from what we don't.

What we know: the official data describes a slowing economy — 1.5% in Q2, 2.1% in Q1, 0.5% in late 2025 — and the Fed sees 2% as the long-run speed limit. We know the AI capex boom is real and enormous: 1.4% of GDP and climbing toward 3%. We know the funding environment is getting tougher and the revenue question is unanswered. And we know Musk has already revised this exact forecast downward once, from double digits to 4%, without acknowledging it.

What we don't know: whether AI delivers economy-wide productivity gains on a twelve-month timetable, whether hyperscaler returns justify the spending, and whether the bond market will keep financing the buildout at current yields. Nobody knows these things — not Musk, not the Fed, not the brokerages. They are genuinely open questions.

So here is the verdict. Musk's direction is plausible: AI investment is already reshaping American growth, and a forecaster who bets on more of it is betting with the visible evidence. But his number is not. Four percent next year would require a productivity miracle with no peacetime precedent, starting from an economy growing at 1.5%, against a central bank that says 2% is the speed limit. That is not a forecast; it is — his word — a guess.

Take the boom seriously. Take the number with salt. And if growth does double next year, remember who told you it would be 10%.

Sources: TheStreet; ANI, via Hindu BusinessLine; BW Businessworld; U.S. Bureau of Economic Analysis; Federal Reserve September 2026 economic projections.