AI Washing Is the New Greenwashing

ESG gave private markets a preview of exactly how this plays out. A claim gets attractive enough that everyone starts making it. Regulators and allocators eventually build the tools to check whether the claim is true. The firms that were honest the whole time gain an advantage. The firms that weren't get a very public, very expensive correction.

AI washing follows the same arc, and it is already further along than most GPs realize.

The mechanics are almost identical. Greenwashing meant claiming environmental impact without the operational substance to back it up. AI washing means claiming AI capability, whether in sourcing, diligence, or portfolio value creation, without the infrastructure, governance, or measurable outcomes to support the claim. Both exploit the same asymmetry: the person making the claim knows far more about what's actually happening inside the firm than the person evaluating it.

That asymmetry is closing fast. Sophisticated LPs are now asking, directly in due diligence questionnaires, how a GP uses AI, what governance sits around it, what risks have been identified, and what controls exist. One experienced operator in the space summarized the new reality bluntly: a fund that says "we use AI" without proof gets dismantled in five minutes by a sophisticated LP asking the obvious follow-up question. That is not a hypothetical. It is happening in fundraising rooms right now.

The regulatory layer is arriving on a similar timeline to ESG's. The EU AI Act classifies credit decisions, employment screening, and financial risk assessment as high-risk uses, carrying compliance obligations and penalties of up to €15 million or 3% of global turnover. Those obligations were set to take full effect on August 2, 2026; a provisional agreement under the EU's "Digital Omnibus" would defer the deadline to December 2027, though that delay is not yet finalized. Either way, funds with European LP bases or European portfolio exposure no longer have the option of treating AI governance as a marketing choice.

There is a specific pattern worth watching for, because it is the AI-era equivalent of a sustainability report with no underlying data. A firm that talks about AI in generic terms, "we leverage AI across the investment process," with no specific answer to what the tool has learned, what it protects against, or what decision it changed, is making a claim it cannot currently defend. A firm that can point to a concrete instance, a risk flagged before it became a write-down, a diligence cycle measurably faster with no drop in quality, a documented change in an IC recommendation, is making a claim grounded in something real.

The lesson from ESG is not that sustainability claims were wrong to make. It's that unsubstantiated claims eventually get tested, publicly, by someone with more information than the person testing them expected to have. AI claims in private markets are entering that same window right now. The firms that build the documentation and the results before the test arrives will not need to worry about it. The firms that don't should start worrying today.