Why is "AI washing" a concern, and what does it mean?

 

This year, revelations that questioned the "Just Walk Out" technology that many of Amazon's real grocery stores had placed garnered negative press for the company.

 

At several of its Amazon Fresh and Amazon Go stores, shoppers can just choose their products and then leave thanks to the AI-powered system.

To determine what you have selected, the AI consults a multitude of sensors. Then, a bill is automatically sent to you.

But in April, it was publicly revealed that Just Walk Out required almost 1,000 staff in India to manually review nearly 75% of the transactions, as opposed to relying just on AI.

Amazon swiftly refuted the accusations, stating that employees in India were not watching CCTV footage from every store and that they were "erroneous."

Rather, it claimed that the Indian laborers were only going over the system. Amazon went on to say, "This is no different from any other AI system where human reviewers are common and that places a high value on accuracy."

Regardless of the specifics of the Amazon case, it serves as a prominent illustration of a recent and expanding inquiry into whether businesses are exaggerating the extent of their usage of artificial intelligence. Referred described as "AI washing" in relation to the environmental phenomena known as "green washing,"

 

Let's review the definition of artificial intelligence first.

 

Artificial Intelligence (AI), which is difficult to define, lets machines learn and solve problems. AI is capable of this after undergoing intensive training on enormous amounts of data.

 

The specific subset of artificial intelligence referred to as "generative AI" has gained a lot of interest lately.

 

When it comes to creating original content, such as text conversations, artwork, and music, this AI shines.

The popular chatbots Gemini from Google, Copilot from Microsoft, and ChatGPT are instances of generative artificial intelligence.

Different kinds of artificial intelligence are used in washing. While some companies make misleading claims about the use of AI while really employing less complicated computation, others exaggerate the effectiveness of their AI in contrast to earlier methodologies or incorrectly claim that their AI solutions are fully operational.

 

Some businesses, in the meantime, are only adding an AI chatbot to their already-existing operational software; this approach does not use AI. According to Open Ocean, an investment business for early-stage internet firms located in Finland and the UK, just 10% of digital start-ups in 2022 revealed utilizing AI in their presentations, but over 25% of them did so in 2023.

 

 

 This year, it anticipates that percentage to exceed one-third.

Furthermore, according to Sri Ayangar, a member of the Open Ocean team, some of these businesses have exaggerated their AI capabilities in an effort to get investment or to look innovative.

"Some founders seem to believe that if they don’t mention AI in their pitch, this may put them at a disadvantage, regardless of the role it plays in their solution," adds Mr. Ayangar.

"And from our analysis, a significant disparity exists between companies claiming AI capabilities, and those demonstrating tangible AI-driven results."

 

Based on statistics from another digital investment firm, MMC Ventures, it's a problem that has been going on for a while. According to a 2019 survey, 40% of brand-new digital companies that referred to themselves as "AI start-ups" really employed hardly no AI at all.

 

"Today's problem is the same problem plus a new one," explains MMC Ventures general partner Simon Menashy.

He said that every business may now get "cutting-edge AI capabilities" for the same price as regular software. Nonetheless, he claims that many businesses are only adding a chatbot interface to a non-AI product rather than developing an entire AI system.

According to Douglas Dick, UK director of emerging technology risk at global accounting firm KPMG, the lack of a single, accepted definition of AI exacerbates the issue of AI washing.

"Everyone in the room would have a different definition of artificial intelligence," he claims. "The phrase lacks a clear point of reference and is used extremely loosely and generically. This uncertainty is what's facilitating the emergence of AI washing.Businesses may experience unsettling effects from AI washing, including as overspending on services and technology or falling short of the operational goals AI was supposed to assist in achieving."

Investors may find it more difficult to recognize truly innovative firms as a result.

 

And according to Mr. Ayangar: "If consumers have unmet expectations from products that claim to offer advanced AI-driven solutions, this can erode trust in start-ups that are doing genuinely ground-breaking work."

At least in the US, regulators are beginning to pay attention. The US Securities and Exchange Commission (SEC) said earlier this year that it was bringing charges against two investment advising companies for making false and deceptive claims regarding the scope of their usage of artificial intelligence.

According to Nick White, partner at international law firm Charles Russell Speechlys, "the firm stance taken by the SEC demonstrates a lack of leeway when it comes to AI washing, indicating that, at least in the US, we can expect more fines and sanctions down the line for those who violate the regulations."

 

The UK currently has rules and regulations addressing AI washing, one of which is the Advertising Standards Authority's (ASA) code of conduct, which requires marketing materials to intentionally mislead or be likely to do so.

 

 

Ads under investigation by the ASA are increasingly using AI claims, according to Michael Cordeaux, the regulatory team associate at Walker Morris, a UK corporate law firm.

As to the ASA, a sponsored Instagram post titled "Enhance your Photos with AI" is an instance of a misleading marketing that exaggerates the program's capabilities.

 

"What is clear is that AI claims are becoming increasingly prevalent and, presumably, effective at piquing consumer interest," says Cordeaux.

 

 

Sandra Wachter, a renowned expert on artificial intelligence from across the world and a professor of technology and regulation at Oxford University, states that "I think we are at the peak of the AI hype cycle."

But I think we've failed to consider whether using AI for every work is necessarily a good idea. I recall seeing advertising for AI-powered electric toothbrushes in the London Tube. For whom is this intended? Who does this benefit?"

She adds that the effects of AI on the environment are frequently downplayed.

 

"Artificial intelligence is not a tree... the technology already has a greater impact on climate change than flying."

 

Instead of incorporating AI into everything, we should start thinking about the specific sectors and tasks that it can assist with, rather than continuing this one-sided, exaggerated argument.

However, Advika Jalan, head of research at MMC Ventures, believes that AI washing may eventually go away on its own.

"AI is becoming so ubiquitous - even if they’re just ChatGPT wrappers - that 'AI-powered' as a branding tool will likely cease to be a differentiator after some time," she asserts. "It will be a bit like saying 'we’re on the internet'."