From ‘cloud washing’ to ‘AI washing’

Remember cloud washing from a few years ago? This term refers to the practice of companies representing their products and services as cloud-based when they are not truly cloud-based. In other words, cloud washing was, and still is, a marketing tactic that capitalizes on the hype surrounding cloud computing without delivering the benefits that true cloud solutions offer.

Today we’re facing the same issues with an even older technology: artificial intelligence. The hype around generative AI is driving this, but AI has been rising as a hyped technology for several years, even though it started in the 1950s. Companies are “AI washing” any technology that’s being sold, no matter if it uses AI or not.

Like cloud washing, AI washing promotes products or services as being “powered by AI,” even though the level of AI integration may be minimal or nonexistent. The goal of AI washing is to make the products or services appear more advanced than they are, for obvious purposes.

For example, a company might label a product as “AI-powered” simply because it uses a basic algorithm that could be considered a form of AI. AI terminology might appear in marketing materials without explaining how AI is used in the product. This can confuse customers who believe they are getting a more advanced or sophisticated product. I’m seeing claims of AI in system management technologies, databases, middleware, security, and pretty much anything that’s being sold today.

AI washing hurts the reputation of the AI industry by creating false expectations and misconceptions about what AI can deliver. It also makes it harder for customers to identify genuinely innovative and useful AI products and services amid a sea of AI-washed offerings.

It’s more difficult to spot than cloud washing, given that you’re taking the word of the salespeople that their technology is indeed “AI-powered.” The technology could make a few simple calls out to a generative AI API and thus, technically, be “AI-powered,” I guess. However, if it’s for no true, useful reason, then it’s just a check mark.

Most of the products and services claiming to have AI capabilities are often underwhelming in terms of any advantages you can see from their use of an AI-driven subsystem. The more honest product managers will admit that their use of AI is anecdotal and not core to the product’s functionality. Of course, the vendor will have plans to leverage more AI capabilities moving forward.

Most technologies, including cloud services, that don’t leverage AI today can’t turn on a dime to suddenly incorporate AI in useful ways. It will take years of product planning and finding the best path for AI within their technology stack before it becomes truly beneficial.

For many of these technologies and cloud services, AI will be more of a forced fit, with no real improvement of the core functionality of the product. Buyers will waste money moving in a technical direction that the market thinks is cool, but without a real requirement to do so, they’re just following the hype.

When I was a product CTO for many years for many companies, I pushed back on these hype-driven fool’s errands, which often put my job at risk. If there is no reason to use a specific technology, including AI, then don’t use it. It only adds complexity, risk, and cost. I’ll die on that hill, and I’m sure it has hurt my career to take that kind of stand.

However, others will bend to the whims of the market. If AI is a hyped topic, then suddenly we sell AI-powered technology, no matter if we should or if it truly improves the core functionality and value of the technology at hand. Therein lies the problem—again.