Aswath Damodaran is sounding the alarm on private credit at the exact moment AI is juicing demand for data centers. In comments carried by Yahoo Finance, the professor of valuation says private credit is “setting itself up for a beating” as the AI data center lending boom accelerates. The gist is not subtle: when a popular theme pulls capital into a tight set of bets, valuation discipline becomes harder, and losses tend to show up later, in a hurry.

Why does this matter right now? Because private credit is typically where investors go when they want yield without the day-to-day turbulence of public markets. But Damodaran’s warning implies a specific failure mode: underwriting that looks fine while growth assumptions are strong can deteriorate when reality lags, when refinancing windows close, or when cash flows underperform the story used to justify the lending. In other words, the “beating” is likely not about one bad loan. It is about how incentives, leverage, and borrower concentration can stack into a broader drawdown when conditions turn.

To understand the stakes, it helps to look at what the AI data center boom is doing to capital markets. Data centers are capital-intensive, long duration projects, and they often require heavy financing over multi-year timelines. When the AI wave drives demand for compute, the financing narrative can tilt toward “mission critical” and “structural growth,” which can make credit terms feel safer than they are. For private credit funds, the pitch usually centers on contracted demand, sponsor quality, and collateral that can be “worked out” if things go sideways. Damodaran’s point is that those comfort blankets can become misleading when everybody is relying on the same macro assumptions at once.

This is where valuation enters. Damodaran is widely associated with valuation frameworks and the discipline of thinking in terms of risk-adjusted outcomes rather than wishful trajectories. When capital floods into a theme, the market can compress spreads, loosen covenants, or accept thinner buffers. Even if the loans are technically “secured,” repayment still depends on operating performance. If the AI-driven growth that lenders underwrite does not materialize at the level expected, or if costs run higher, debt service coverage can fall quickly. Private credit is not immune to that dynamic. In fact, because the deals are less visible than public bond markets, problems can be slower to surface, then more abrupt when they do.

There is also a governance angle for decision-makers. Private credit often involves committees, deal-by-deal negotiations, and internal debates over what counts as acceptable risk. When a strategy becomes popular, the social pressure to participate rises, and the internal bar can drift. Boards and investment committees may find themselves comparing “relative performance” against peers in similar sleeves, especially if fundraising depends on deploying capital. Damodaran’s warning can be read as a reminder that outperforming a cycle does not prove that risk was priced correctly. It can mean the market was generous, and the generosity can reverse.

Regulation does not necessarily prevent this specific kind of risk, either. Private credit markets operate under a different set of disclosure and oversight dynamics than public markets. That does not mean rules are absent; it means information moves differently, and transparency can lag behind the speed of capital allocation. When losses occur, investors and regulators often shift from “monitoring” to “questioning how underwriting assumptions got so optimistic.” The cycle can then become not just a credit issue, but a reputational and structural one.

So what is the strategic takeaway for executives facing similar themes? The lesson embedded in Damodaran’s warning is about concentration and comfort. If a portfolio is tilted toward AI data center lending, it is worth stress-testing the parts of the underwriting that are most theme-dependent: occupancy and demand projections, cost inflation, refinancing assumptions, and the liquidity of collateral in a stress scenario. Even without changing the thesis, boards can push for clearer scenario analysis tied to cash flow timing, covenant headroom, and sponsor leverage. The aim is to avoid getting “beaten” by a mismatch between expected and actual timing of cash generation.

If AI continues to drive infrastructure demand, private credit could still earn attractive returns. But Damodaran’s point is that the downside is not hypothetical when everybody is leaning the same way. For investors, lenders, and boards underwriting in AI-linked corners of the economy, the priority is not chasing the narrative. It is pricing the risk that narratives compress during good times, and that the beating arrives when the cycle stops cooperating.