Shortly after the release of OpenAI's ChatGPT in 2022, news about generative AI exploded, generating huge interest and sparking a hype cycle that peaks with each new model or feature. Suddenly, it seems like every tech company and startup is selling an AI product that promises to revolutionize consumer behavior.
A survey by an investment fund for emerging technology companies found that the number of startups mentioning AI in their presentations rose from 10% in 2022 to more than 25% in 2023. More than 50% of S&P 500 companies mentioned AI in their earnings calls last year, the investment fund's report said. NBC News.
But how many of these companies are actually doing it? A survey conducted by the U.S. Census Bureau last November found that only 4.4% of U.S. companies use AI in the production of goods and services. A 2019 survey by London-based venture capital firm MMC found that 40% of European AI startups don't use AI at all.
At the root of “AI washing” is technology companies and startups advertising that they are using AI, but in reality they are not.
What is AI cleaning?
AI washing is a term derived from greenwashing, where companies make exaggerated claims about their environmental friendliness to customers. Similarly, companies that claim to have integrated AI into their products may be accused of AI washing, when in fact they use less advanced technology.
Advertisements that exaggerate the capabilities of a company's AI tools or mislead consumers about features that aren't yet operational in AI products also qualify as AI washing.
It is unclear who coined the term “AI washing,” but the term was popularized in March 2024 when the US Securities and Exchange Commission (SEC) fined investment advisory firms Global Predictions and Delphia $225,000 (Rs 180 crore) and $175,000 (Rs 140 crore), respectively. The securities regulator found that the companies had made false representations to clients about offering “expert AI-driven predictions” and using machine learning to manage private client portfolios.
What are some examples of AI cleaning in practice?
The rapid advancement and huge potential of AI has forced many companies, including several tech giants, to cut corners when it comes to rolling out AI-based products. For example, Google launched its flagship multimodal AI chatbot, Gemini, last year with a demo video showcasing its ability to recognize images and real-world objects. In one part of the video, a person draws a picture of an animal and Gemini correctly guesses it to be a duck.
However, it turns out that the video was not shot in real time. In fact, Google Bloomberg The video was reportedly created by sending text prompts to Gemini and stitching together still images. A disclaimer was included in the YouTube clip's description, but there was no such disclosure in the video itself.
Recently, Amazon has reportedly removed cashierless checkout systems from many of its grocery stores. Business Insider “Just Walk Out” technology, which claimed to use AI and sensors to detect the contents of customers' shopping carts, was found to actually rely on employees in India to verify transactions.
AI-related claims by not only tech companies but also multinational brands such as McDonald's and Coca-Cola have come under scrutiny. McDonald's recently decided to remove AI technology from its US drive-thru restaurants after customers complained that their orders were being cancelled by mistake. Meanwhile, Coca-Cola tried to jump on the AI bandwagon by introducing a limited edition flavour of AI-generated iced drinks last year, but ultimately failed to impress many customers.
There are also an increasing number of AI apps boasting chatbot capabilities, but in reality they are just ChatGPT wrappers, meaning the underlying technology powering the app is not theirs but OpenAI's.
In India, Ola founder Bhavesh Agarwal's startup released a beta version of Krutim AI, touted as a home-grown ChatGPT rival. However, after the chatbot confirmed that it was “created by OpenAI,” many users began questioning whether Krutim AI was a ChatGPT wrapper. In response to the speculation, the startup said it had investigated the issue and found that the root cause was “a data leakage issue due to one of the open-source datasets used in the fine-tuning process of our Large Language Models (LLM).”
The electronics manufacturing industry has not been immune to the effects of AI cleaning. For example, French company Kolibree introduced an AI-powered toothbrush in 2017, which essentially uses sensors to detect how users brush their teeth, and a “deep learning algorithm” analyzes that data to provide personalized recommendations for users to improve their brushing. According to reports, the annual consumer trade fair CES 2024 in Las Vegas, USA, was packed with law graduates and saw a surge in gadgets sold under the banner of generative AI.
The rush to brand itself as an AI business follows a longstanding pattern of companies looking to cash in on the hype surrounding emerging technologies. At the peak of the cryptocurrency boom in 2017, Long Island Iced Tea Company rebranded to Long Blockchain Corp., but in 2021 the U.S. Securities and Exchange Commission deregistered the beverage maker after it was reported that it never operated a blockchain business.
Why are there growing concerns about AI washing?
AI washing may seem like a harmless hyperbole, but it could have far-reaching implications for consumers and the tech industry itself.
Linda Yao, vice president of AI Solutions and Services at Lenovo, explained: ZDNet AI washing diverts executive attention and resources from practical AI innovation. Instead of developing meaningful AI capabilities, companies may misinvest in cosmetic enhancements, slowing real progress in the technology.
Yao also noted that AI washing can complicate decision-making for businesses that are truly seeking valuable AI solutions. This can hinder companies' digital transformation efforts, stifle innovation and jeopardize financial performance, he said. Regarding consumers, Lenovo's senior executives said substandard AI technology can pose data security and privacy risks, further alienating consumers from using the technology.
To avoid (even unintentionally) AI washing, the US Federal Trade Commission (FTC) recommends that companies rein themselves in by asking key questions: “Are you exaggerating the capabilities of your AI product? Are you promising that your AI product is better than a non-AI product? Does your product actually use AI?”
“Before you label a product as AI-powered, keep in mind that simply using AI tools in the development process is not the same as having AI built into the product,” the FTC added.
Meanwhile, the Securities and Exchange Board of India (SEBI) had issued some warning against AI washing in a 2019 circular.
“Since most AI/ML systems are black boxes and their operation cannot be easily quantified, it is essential to ensure that any financial benefits from these technologies are not misrepresented when advertised in investor financial products offered by intermediaries,” the notice states.
