Alright, there’s a serious issue we need to talk about. And, surprise! It involves artificial intelligence, specifically, something called AI washing. It seems a bit strange: People went from skeptical about AI to eager, and now brands are attempting to capitalize on the hype by exaggerating how they use it or what their tools are capable of.
What businesses may not realize is that their actions have consequences — and some brands are beginning to find themselves in hot water.
So, here are the questions I want to answer:
- What does AI washing actually mean?
- What are the consequences of AI washing?
- Who’s guilty of AI washing?
- What’s the right way to approach AI in marketing messaging?
What Is AI Washing?
The term AI washing refers to overstating, exaggerating or misrepresenting the use of artificial intelligence in products, services or business operations. It’s pretty much that simple. When brands engage in AI washing, their ultimate goal is typically to appear more innovative and competitive in a world where AI adoption is growing fast.
Here are a handful of typical ‘tactics’ brands who engage in AI washing might use:
- Buzzword Exploitation: Using “AI” as a buzzword without substantiating its actual application or impact.
- Minimal AI Implementation: Highlighting AI features when the product uses basic algorithms, simple automation or non-AI technologies instead.
- Misleading Claims: Giving the impression of advanced capabilities to attract investment, customers or media attention.
In the same vein as making misleading claims about AI’s capabilities to perform a task, understating or simply failing to mention just how much human involvement is still required in an AI process to produce acceptable outcomes counts as AI washing.
For example, you shouldn’t tout your AI tools as fully autonomous if you still need humans to help out on the backend. Even with something as little as quality assurance, humans are still involved, so your claim of full autonomy could be misinterpreted.
The problem may be obvious, but it also may be true that businesses just don’t believe that exaggerating the truth is that big of a deal. In reality, it’s no better than lying, which isn’t OK to do in marketing messages.
The Implications of AI Washing
Before we explore some real-world examples of AI washing and how it’s impacted some brands, let’s consider some general consequences that could come of this deceptive practice:
Erosion of Customer Trust and Reputational Damage
Algorithms and paid influencers have already begun to erode trust among consumers, who are increasingly seeking real-world experiences to inform their decision-making. Regarding sharing real experiences, Reddit is one of the foremost platforms on which consumers do this. Direct recommendations from other users prove far more valuable than generic reviews for driving purchase decisions.
But to get those recs, businesses have to not only offer a solid product or service but be trustworthy.
AI washing wasn’t the thing that catalyzed this erosion of trust in the first place, but it certainly could be the one that breaks the camel’s back if brands aren’t careful. Once consumers realize that a brand’s AI claims are exaggerated or false, it can lead to an accelerated loss of trust, negative media coverage and eventually criticism that consumers will inevitably amplify on social media. Once a business has tarnished its reputation, it can be incredibly difficult — and expensive — to earn back.
Financial Loss
Businesses lose money in tons of different ways, but negative backlash from AI washing coming to light poses various financial risks, such as:
- Lost revenue as consumers stop buying.
- The withdrawal of funding by investors.
- Increased marketing costs via damage control efforts, rebranding or rebuilding to regain lost trust.
Legal Risks
In some jurisdictions, AI washing could be considered misleading advertising, subjecting brands to fines. As I’ll showcase a bit later, some companies are already facing hefty penalties. But more than that, misleading or false AI claims in a bid to attract investors could lead a business right into a fraud lawsuit, which often carries far higher stakes for revenue, reputation and recoverability.
Real-Life Examples of AI Washing
We’ve already seen some action when it comes to AI washing, including big names being called out en masse to expensive lawsuits. Have a look:
Coca-Cola Y3000 Soda
You may be familiar with this one, but if you aren’t here’s a brief refresher:
Last year, Coca-Cola debuted a new flavor of soda called Y3000, claiming it was “co-created” with AI. Whether that claim was true or not doesn’t necessarily matter because they failed to provide further details about how they utilized AI in the product creation process.
This left critics questioning the extent of AI use or its involvement in the process at all. Could it have just been a ploy to grab consumer attention? That’s a likely answer.
U.S. Investment Firms
Earlier this year, the U.S. Securities and Exchange Commission (SEC) fined a couple of investment firms for making misleading AI claims. They both eventually settled out of court, ordered to pay fines of up to $400,000.
A quote from the press release claims that both companies “marketed to their clients and prospective clients that they were using AI in certain ways when, in fact, they were not.” So, in this case, the AI washing was more lying than it was exaggerating capabilities — but both are no good.
Put a Kabash on AI Wash
I talk a lot on this blog about being honest and ethical when it comes to AI, which mostly comes across by way of messaging like “if you’re using AI, say that you’re using it,” or similar. But for this discussion, I want to look at the other side of the honesty coin: Being careful not to overstate AI’s capabilities in your disclaimers or claims.
Yes, you absolutely should be transparent about AI usage in your products or services, but take the time to craft disclaimers that are fair. Here are some tips to help you do just that:
Create Messaging That Explains What AI Does and Doesn’t Do
It seems easy to write something that explains how your AI tools work, but it can be just as easy to get excited and carried away — which could lead to AI washing. When creating your AI messaging, consider writing both what it can and cannot do and acknowledging its limitations. This helps paint a clearer picture for consumers of what the tool is truly capable of.
Back AI Claims with Hard Evidence
One of the best things you can do to prove your claims is to back them up with evidence. If you say that your AI is capable of something, show it, whether through a case study with measurable results, a real, live-action video demonstration, etc. At any rate, you should always be prepared to answer questions about how your AI works in practice, even if you don’t share proprietary details. If something seems too good to be true, people will start sniffing around, and it’s only a matter of time before consumers or authority figures find out.
Seek Testimonials and Third-Party Validation
You can say anything you want, but these days, most consumers take company claims at face value. If you offer an AI tool, share testimonials from users or credible sources that validate its capabilities. To that end, encourage neutral third-party reviews of your AI technology to build credibility.
You Know the Drill
AI is difficult and in a weird place right now. It’s neither a new thing nor a well-established technology. A lot of folks are still trying to find their footing when it comes to artificial intelligence, and having to make claims about how, when and why you’re using it adds a level of complexity to an already confusing game for some.
Lying is never OK — we know that. But I also understand that it’s easy to get excited about new technology, which could lead to inadvertent AI washing. If you feel like you’re susceptible to going down that path, take a step back and review your messaging. Are you telling consumers that AI is involved in your products, services or processes? What kind of language are you using? Can you back up your claims with evidence?
Ask yourself these questions and others like them to develop a deeper understanding of AI’s role in your business. Once you nail it down, craft your messaging accordingly.
