David Solomon runs Goldman Sachs, a bank positioned at the very center of financing the AI boom. So when he sits down at The Economic Club of Washington and tells you, plainly, how much of that boom is narrative rather than substance, it is worth listening with unusual care. The person selling the boom is describing its mechanics.
Solomon’s framing was precise and, for someone in his position, remarkably candid. AI, he argued, has re-empowered the head of technology in every American business and pushed budgetary resources into the hands of the chief technology officer. But the thing actually driving the economic impact, he said, is American businesses spending more on digitization broadly — of which AI is just one component. AI, in his words, is in some sense the cover story that creates the space for American companies to embrace technology far more profoundly than they otherwise would.
That word — cover story — from the CEO of Goldman Sachs, is the most important thing a leader can understand about the financial reality of AI in 2026. This piece unpacks what he meant and what it means for your organization’s money.
The $500 billion question
Solomon put a number on the scale of the boom that makes the stakes concrete. Data-center spend in the United States this year, he noted, is over half a trillion dollars — over $500 billion, in a single year. That figure is corroborated by outside analysis: McKinsey and S&P Global estimate that adding even 50 gigawatts of new data-center capacity by 2030 requires more than $500 billion in internal data-center infrastructure alone, with hyperscalers planning $300 billion-plus in 2025 by themselves. Goldman’s own research forecasts figures rising toward $5 trillion by 2030.
Then Solomon said the quiet part directly. You are not going to generate this kind of spend, he observed, unless you are going to make a promise you are going to profoundly change the world. Is it hype? Of course, he said. And then the line every leader should tape to their monitor: How else are you going to get people to write $500 billion of checks just this year alone? There needs to be a level of, like, AI is your savior, almost.
This is a sitting bank CEO explaining that the scale of the required capital demands a proportionally enormous promise. The world-changing narrative is not merely enthusiasm. It is the financial precondition for the spend. You cannot raise half a trillion dollars a year on a modest, measured, realistic projection. You can only raise it on salvation.
Why the “cover story” framing matters for your budget
Solomon’s insight is not that AI is worthless — he was explicit that AI will profoundly change the world, and that in areas like call centers and software engineering the productivity gains are real. His insight is subtler and more useful: the AI narrative is doing double duty. It is both a genuine bet on a transformative technology and a cover story that unlocks broader digitization budgets that would otherwise be hard to justify.
This matters enormously for how you read your own organization’s AI spending. Some of what is being funded under the AI banner is genuine AI investment with a real return thesis. Some of it is broader digitization — modernization, cloud migration, data infrastructure — that is valuable but is being labeled AI because AI is the narrative that gets the check signed. And some of it is neither: spending justified purely by the fear of missing the salvation, with no return thesis at all.
The disciplined leader separates these three. The digitization dressed as AI may be a perfectly good investment — but it should be evaluated as digitization, on digitization’s returns, not waved through on AI’s salvation narrative. And the spending justified only by fear should be identified and stopped, because it is the purest form of paying for the cover story itself.
The bank’s own research undercuts the boom
The most striking corroboration of Solomon’s skepticism comes from Goldman Sachs itself. In March 2026, Goldman’s own research found no meaningful relationship between AI adoption and productivity gains at the economy-wide level — even as 70% of S&P 500 management teams were discussing AI on their earnings calls. Seventy percent of the largest companies in America are talking about AI; the measurable economy-wide productivity link is, according to the bank at the center of the boom, not yet there.
This aligns with the MIT Media Lab finding from July 2025 that 95% of organizations saw no measurable return on their generative-AI investments despite billions in spending. Two of the most credible institutions in the space — MIT and Goldman itself — are independently reporting that the returns, so far, are not showing up in the aggregate numbers, even as the spending and the talk reach record highs. The gap between the narrative and the measured return is not a fringe claim. It is the finding of the establishment.
Solomon’s own dinner table anecdote
Solomon illustrated the gap with a story. He was in China with a group of prominent global executives, all talking about how AI was changing their business. He proposed they go around the table and each share a story of how they were using AI to make their business better. He heard five or six great stories — and not one of them involved generative AI. The improvements were digitization, automation, data — the broader category. The generative AI that dominates the narrative was absent from the actual value stories.
This is the cover story made visible. The executives were, in their own minds, telling AI stories, because AI is the frame everyone now reasons inside. But the actual value came from the broader digitization that AI’s narrative had made room for. The narrative and the substance had quietly separated, and only a direct question — tell me the specific story — revealed the gap.
Mapped to the Mutation Readiness framework
Solomon’s financial framing maps onto three dimensions of the Mutation Readiness diagnostic — the operational instrument of the Mutation transformation practice we run for enterprise leaders.
Signal Sensitivity — Solomon’s dinner-table move is a Signal Sensitivity masterclass: he did not accept the stated signal (everyone says AI is transforming their business) and instead read the measured signal (name the specific story), which revealed the gap. A leader with high Signal Sensitivity does the same with their own budget — distinguishing the AI narrative from the digitization substance from the fear-driven waste, rather than accepting that all spending under the AI banner is AI value.
Ambidextrous Capital — the entire cover-story dynamic is a test of this dimension. Ambidextrous Capital requires knowing precisely what each dollar is buying: proven exploit value, genuine explore bet, or narrative participation. Solomon is describing an environment where the three have been deliberately blurred because the blur is what unlocks the budget. The disciplined organization un-blurs them, funding the digitization as digitization and the AI bets as measured explore, and refusing to fund the salvation narrative itself.
Ethical Guardrails — framed as containment-as-velocity. The $500 billion boom is velocity at civilizational scale, driven by a salvation narrative. The guardrail is the discipline that keeps your own organization’s spending contained within measured-value bounds rather than swept along by the narrative that, as Solomon admits, exists precisely to get people to write checks they otherwise wouldn’t.
The signals your organization is missing right now
The master signal is how much of your AI budget is buying AI, how much is buying digitization wearing an AI label, and how much is buying nothing but participation in the narrative. Most organizations have never separated the three, which means they cannot see how much they are paying for the cover story itself.
Look for the specific tells. If you asked your leaders to name the specific value story behind each AI investment — as Solomon asked the executives in China — how many could name a real one, and how many would name digitization or nothing? Which of your AI-labeled projects are actually broader modernization that should be evaluated on its own returns? Where is spending being justified by the fear of falling behind rather than by a return thesis? Each is a place the cover story is spending your money.
Three practical questions
One: for each AI investment, can your team name the specific value story — the way Solomon demanded at that dinner? If the honest answer is digitization, or a vague gesture at transformation, you have found spending that is riding the cover story. Re-evaluate it on what it actually delivers, not on the narrative that got it approved.
Two: how much of your “AI budget” is really digitization? Broad digitization may be an excellent investment — but it should be justified as digitization, on its own returns, not approved under an AI salvation narrative that suppresses scrutiny. Separate the two, and hold each to its own standard.
Three: what are you spending purely because AI is “your savior”? Solomon named the mechanism: the narrative exists to get checks written. Find the spending in your organization that has no return thesis beyond fear of missing the salvation, and stop it. That spending is the cover story cashing your check.
The closing thought
When the CEO of Goldman Sachs — an institution positioned to profit enormously from the AI boom — tells you that AI is partly a cover story, that the $500 billion in annual spend requires a salvation narrative to justify it, and that his own bank’s research finds no economy-wide productivity link yet, he is handing you an extraordinary gift. He is showing you the mechanics of the boom from the inside.
None of this means AI is fake or that you should stop investing. Solomon believes, and the evidence supports, that AI will profoundly change the world and that in specific domains the gains are already real. It means something more precise and more useful: the narrative and the substance have separated, deliberately, because the scale of the required capital demands a promise proportional to salvation. The disciplined leader reads the money the way Solomon reads it — separating the genuine AI bet from the digitization wearing its label from the pure narrative participation — and funds only the first two, on their own honest returns.
The people writing $500 billion of checks need AI to be salvation. You do not. You need it to deliver a measured return. Hold it to that standard, and you will spend on the substance while everyone else pays for the cover story.
The world has changed. The leaders who notice will be the ones the next decade is built around.
