AI Washing: How Companies Exaggerate Tech Capabilities

Artificial Intelligence is the hottest buzzword in the corporate world right now. From Wall Street earnings calls to startup pitch decks, executives are racing to tell you how their “proprietary algorithms” will revolutionize the industry. But there is a growing problem lurking beneath the hype. Much like “greenwashing” involves exaggerating environmental efforts, “AI washing” occurs when businesses fabricate or overstate their use of artificial intelligence. Regulators are now stepping in to protect investors and consumers from these misleading claims.

What Is AI Washing?

AI washing is a deceptive marketing practice where a company promotes its products or services as being powered by artificial intelligence when the technology involved is actually quite basic—or nonexistent.

In some cases, companies rebrand simple data analytics or rule-based software as “advanced machine learning.” In more severe cases, businesses claim to use sophisticated neural networks to predict stock prices or manage tasks, while the work is actually being done by low-paid human workers or standard spreadsheets.

The goal is usually financial. Companies that attach “AI” to their name or product description often see higher stock valuations and increased interest from venture capital investors.

The SEC Crackdown: Fines and Penalties

The United States Securities and Exchange Commission (SEC) has made it clear that they are watching. In 2024, the agency moved from issuing warnings to issuing fines. SEC Chair Gary Gensler has explicitly stated that public companies must ensure their disclosures about AI are accurate and not misleading.

The Delphia and Global Predictions Cases

In March 2024, the SEC announced settled charges against two investment advisers for making false and misleading statements about their use of AI. This marked the first major enforcement action specifically targeting AI washing.

1. Delphia (USA) Inc.: This Toronto-based firm claimed its AI could predict which companies and trends were about to grow, allowing them to beat the market. They marketed their ability to use client data to feed a “predictive algorithmic model.” The SEC investigation found these capabilities did not exist. Delphia agreed to pay a civil penalty of $225,000.
2. Global Predictions Inc.: Based in San Francisco, this company marketed itself as the “first regulated AI financial advisor.” They claimed to offer “expert AI-driven forecasts.” The SEC found that these claims were false. The firm also included hypothetical performance data on its website without disclosing that it wasn’t real trading history. Global Predictions agreed to pay a civil penalty of $175,000.

These fines serve as a concrete warning to other financial firms: if you claim you are using AI to manage money, you must be able to prove it.

The FTC and "Operation AI Comply"

The Federal Trade Commission (FTC) is also aggressively targeting companies that deceive consumers with fake AI claims. In September 2024, the FTC announced “Operation AI Comply,” a law enforcement sweep against five companies.

One of the most high-profile targets was DoNotPay, a company that billed itself as “the world’s first robot lawyer.” The service claimed it could generate legal documents and even sue corporations on a consumer’s behalf using artificial intelligence. The FTC complaint alleged that the company did not actually conduct the quality of legal analysis it promised. DoNotPay agreed to pay $193,000 to settle the charges.

Other companies targeted in this sweep included:

• Ascend E-com: Claimed their AI tools would help consumers earn thousands of dollars in passive income through online stores.
• Ecommerce Empire Builders: Promised students they could build “AI-powered” empires.

The "Human in the Loop" Controversy

A common form of AI washing involves hiding human labor behind a digital curtain. This is sometimes called the “Wizard of Oz” technique. Companies present a seamless, automated interface to the customer, but the heavy lifting is done by remote contractors.

A notable example involves Amazon’s “Just Walk Out” technology. For years, this system was marketed as a seamless AI experience where computer vision tracked what you took off the shelf so you could leave the store without checking out. However, reports in early 2024 revealed that the system relied heavily on over 1,000 workers in India who manually reviewed video footage to ensure transactions were accurate.

While Amazon stated that the human reviewers were primarily there to annotate data and train the model, the revelation damaged the public perception of the technology. It highlighted a reality of the current tech landscape: many “automated” systems are just people working behind the scenes.

How to Spot AI Washing

Investors and consumers need to look past the marketing copy. Here are three signs a company might be exaggerating its tech capabilities:

1. Vague Terminology: Be skeptical of phrases like “proprietary neural engine” or “advanced cognitive computing” if the company cannot explain what the software actually does. If they cannot define the input and the output clearly, it might be marketing fluff.
2. Sudden Pivots: If a company has existed for ten years as a standard software provider and suddenly rebrands as an “AI-first” company without hiring new engineering talent or releasing new products, they are likely just riding the hype cycle.
3. The Wrapper Effect: Many startups today are simply “wrappers” for OpenAI’s GPT-4 or Anthropic’s Claude. While these tools are useful, a company claiming to have built its own massive AI model when it is actually just paying for an API subscription is misleading investors about its intellectual property.

Frequently Asked Questions

What is the difference between AI and an algorithm? An algorithm is a set of rules a computer follows to solve a problem. Simple software uses algorithms (like a calculator). AI, specifically machine learning, involves systems that can learn from data and improve over time without being explicitly programmed for every single outcome.

Why is AI washing illegal? It falls under fraud and consumer protection laws. If a company lies about its capabilities to get you to buy their product or buy their stock, that is a material misrepresentation. The SEC protects investors from financial lies, while the FTC protects consumers from deceptive advertising.

Has anyone gone to jail for AI washing? Currently, most enforcement actions have resulted in civil penalties and fines (like the charges against Delphia and Global Predictions). However, if the fraud is massive and intentional (like the Theranos scandal, but for AI), criminal charges could theoretically follow in the future.

Does using ChatGPT count as AI washing? Using ChatGPT is not AI washing. However, claiming you built the technology behind ChatGPT when you are simply renting it constitutes AI washing. Companies must be transparent about whether they own the technology or are licensing it from a third party.