--- language: en title: "AI washing" description: "AI washing means presenting AI as playing a greater or different role than can be substantiated. Commercial claims must match its demonstrated operation." canonical: "https://tropk.ai/en/resources/glossary/ai-washing" last_modified: "2026-10-12T16:09:29.465Z" --- # AI washing Reviewed on 2026-10-07. ## What it means AI washing is the misleading or exaggerated presentation of [artificial intelligence](/en/resources/glossary/artificial-intelligence)’s role in a product, service, or process. It can occur when a company claims to use AI at a stage where it does not operate, or attributes capabilities to the system that it cannot demonstrate. The issue is the gap between the commercial claim and verifiable operation, not simply the use of AI. ## Why it matters for marketing For marketing, checking these claims helps avoid promising a capability different from what the customer will receive. The team needs to identify the assisted stage, the data used, the limitations, and the human involvement. A demonstration of another version or an exceptional case does not automatically support the promise. Describing the process accurately makes it possible to assess the offer without relying on a label. ## Example in practice Imagine a company advertising that its system uses AI to predict inventory replenishment. In the demonstration, the feature merely applies fixed limits entered by a person, without the advertised mechanism. The team requests evidence of the prediction, finds that the promise is unsupported, and revises the feature description. It keeps useful automated rules, described as such, without inventing an additional capability. ## How to apply or verify Compare each commercial statement with a relevant demonstration and product documentation. Specify the surface, version, task, and conditions, distinguishing between a prototype, an available feature, and a human decision. Record what has been substantiated and what remains unresolved. In 2024, the SEC announced cases involving misleading AI claims by investment advisers in the US; that context does not automatically define the obligations of any Brazilian product. ## Important distinctions Rule-based automation is not necessarily AI, but using AI in a small step does not make every claim about it false. Accuracy depends on the scope of the claim. AI washing differs from a model response that invents features without company involvement. Questions about architecture call for investigation, not automatic accusations; insufficient evidence should lead to narrowing or correcting the promise. ## Sources and references - [SEC Charges Two Investment Advisers with Making False and Misleading Statements About Their Use of Artificial Intelligence]() — U.S. Securities and Exchange Commission ## Continue a leitura - [Hallucinated claims about a brand]() - [AI agent]() - [Brand fidelity]() ## About the author [Luiz Gardelin]() Luiz Gardelin is the founder of Tropk.ai and works in GEO/AEO, product vision, strategy, and growth, helping brands understand and improve their presence in AI responses. --- [All pages in Markdown]()