Glossary

AI washing

Written with AI from the sources below and checked against our facts sheet. Last checked 7 Nov 2026. How this site is made · Spot an error?

AI washing is overstating or misrepresenting how much a product uses AI or what its AI can do. Examples include calling a keyword filter "advanced AI" or promising "100% accuracy." In the US it can be deceptive under Section 5 of the FTC Act. The SEC has also fined investment advisers for false AI claims.

In plain English

AI washing is the AI version of greenwashing: the label is bigger than what is inside. It comes in two forms.

Claiming AI that isn't there. A rules-based chatbot sold as "advanced AI."
Overstating real AI. "Perfectly accurate," "removes human bias" or "replaces your lawyer."

Regulators judge the overall impression a claim creates. A footnote saying "results may vary" does not fix a misleading headline.

Example

Quillmark Software, a fictional HR-tech company, drafts a homepage promising that its hiring assistant is "100% bias-free and 98% accurate." The facts behind it:

The product is a keyword matcher plus a small ranking model.
Internal tests showed 84% agreement with experienced recruiters, on one client's data.
Bias testing covered sex but not age or disability.

The governance review removes "100% bias-free," an absolute no one can substantiate. It replaces "98% accurate" with what was measured, on which data. It adds that recruiters review every shortlist, and it keeps the test evidence on file before the page goes live.

Why it matters

FTC. Under Section 5, a claim is deceptive if it is likely to mislead a reasonable consumer and is material. Advertisers need a reasonable basis before making a claim. The FTC's Operation AI Comply sweep of 25 Sept 2024 brought five actions. Its chair said there is "no AI exemption from the laws on the books." In December 2025 the FTC set aside its Rytr order, citing the AI Action Plan. It said it will still act against companies that deceive consumers about AI capabilities.

SEC. On 18 Mar 2024 the SEC charged two investment advisers, Delphia and Global Predictions, over false claims about their use of AI.

States. In September 2024 the Texas Attorney General settled with Pieces Technologies over allegedly inaccurate claims about its clinical AI's hallucination rate.

EU. The Unfair Commercial Practices Directive bans misleading actions and misleading omissions toward consumers.

How the AIGP exam tests it

AI washing appears under consumer protection (lesson II.B.4) and external communication (lesson IV.C.6).

Common traps:

Deception vs unfairness. AI washing is deception: what you say. Unfairness is what you do, such as deploying untested facial recognition.
Fine print. A small disclaimer cannot cure a prominent false claim.
Puffery. Vague boasting is tolerated. "Perfectly accurate" and "never used for training" are specific, checkable claims.
Fix it later. Claims need evidence before publication, not after a complaint.
"Not our remit." Overstated AI claims are a legal and trust risk, so governance should review them.

The tell: compare the claim with the test evidence. Any gap is the problem.

Frequently asked questions

What is an example of AI washing?

Marketing a simple keyword filter as "powered by advanced AI" is one. Claiming a hiring tool is "100% bias-free" or a detector is "99% accurate" without test evidence is another. In 2024 the SEC found that two investment advisers had made false claims about their use of AI. The firms paid civil penalties totaling $400,000.

Is AI washing illegal?

It can be. In the US, a claim that is likely to mislead a reasonable consumer, and is material, is deceptive. Section 5 of the FTC Act and state laws ban such claims. False AI claims to investors can breach securities rules. In the EU, the Unfair Commercial Practices Directive bans misleading actions and omissions, AI claims included.

What did the SEC do about AI washing?

On 18 March 2024 the SEC settled charges against Delphia (USA) Inc. and Global Predictions Inc. It found they had made false and misleading statements about their use of AI. Delphia paid $225,000 and Global Predictions $175,000. Neither admitted or denied the findings.

How can companies avoid AI washing?

Make claims match the evidence. Describe what the AI actually does, its role alongside people and its known limits. Back every performance number with tests on stated data before publishing it, and avoid absolutes such as "bias-free" or "perfectly accurate." Have governance, legal and communications review AI claims, and update them when the system changes.

Sources

Spot an error? Tell us