Anna Totterdell
Projects Director
Open any enterprise software website right now. Count how many times you see the phrase "AI-powered." It is on every landing page, every product description, every pitch deck. CRMs are AI-powered. ERPs are AI-powered. Project management tools, invoicing platforms, HR systems - all AI-powered.
And in the vast majority of cases, it means absolutely nothing.
"AI-powered" has become the "cloud-based" of 2026. A decade ago, every software product rushed to add "cloud" to its marketing, regardless of whether the architecture had meaningfully changed. Today, the same thing is happening with AI. Vendors are wrapping a language model call around an existing feature, adding the word "AI" to the button, and charging you more for it.
This is not innovation. This is labelling.
What "AI-powered" usually means
In practice, when a software vendor says their product is AI-powered, they typically mean one of three things:
One: they have added a text generation feature. You can now generate email drafts, product descriptions, or summaries using an embedded language model. This is useful but trivial - it is the same capability you already have in ChatGPT, just placed inside a different interface.
Two: they have added a recommendation engine. The software now suggests next actions, predicts outcomes, or scores leads. This can be valuable, but only if the underlying data is clean and the model is trained on relevant information - which, in most implementations, it is not.
Three: they have added a chatbot. You can now ask questions of your data using natural language. This sounds impressive in a demo and is almost universally disappointing in practice, because the quality of the answer depends entirely on the quality of the data it is querying, and most business data is not structured for this kind of interaction.
None of these things are bad. Some of them are genuinely useful. But none of them justify the breathless marketing that surrounds them, and none of them will transform your operations.
The question you should be asking
When a vendor tells you their product is AI-powered, there is only one question that matters: what specific operational outcome does this produce that was not possible before?
Not "what can it do." What does it actually change? Does it reduce a process that took five hours to thirty minutes? Does it catch errors that previously went undetected? Does it enable decisions that previously required a week of manual analysis?
If the answer is specific and measurable, you might be looking at something genuinely useful. If the answer is vague - "it makes you more productive," "it gives you insights," "it helps you work smarter" - you are looking at a marketing feature, not an operational one.
Why this matters for your AI investment
The "AI-powered" inflation is not just annoying. It is actively harmful, because it distorts where businesses invest their AI budgets.
If you believe that buying AI-powered software is the same as implementing AI in your operations, you will spend your money on upgraded licences instead of operational integration. You will tick the "AI" box on your strategy document without actually changing how your business runs. And you will be genuinely surprised when, twelve months later, none of your metrics have moved.
The businesses getting real value from AI are not getting it from their software vendors' AI features. They are getting it from AI enablement designed around their specific processes, their specific data, and their specific pain points. They are getting it from AI that is embedded into workflows, not added to toolbars.
The gap between features and systems
There is a fundamental difference between an AI feature and an AI system. A feature is something a software product offers you. A system is something that is designed around how your business works.
A feature might summarise a customer email. A system - the kind that comes out of proper data and systems integration - routes that email based on content, extracts key data, updates the relevant records in your CRM and ERP, flags exceptions for human review, and triggers the next step in the workflow automatically, reliably, every time.
The feature is impressive in a demo. The system is what changes your operation.
Most vendors sell features. What your business needs is systems. And the gap between the two is where your actual AI opportunity lives.
How to evaluate AI claims
Next time you are evaluating a product, a vendor, or a proposal that uses the word AI, apply this filter:
Does it change a process, or does it add a button? Does it connect to my data and my workflows, or does it operate in isolation? Can I measure the impact in a specific metric, or is the value described in generalities? And most importantly: could I achieve the same outcome with a structured workflow and a well-designed integration, without the AI label?
Often, the answer to that last question is yes. And that should tell you something important about where the real value lies.
It is not in the label. It is not in the feature. It is in the operational design - the boring, structural IT and process strategy work of making systems work together, data flow properly, and processes run without manual intervention.
"AI-powered" means nothing until it is connected to how your business actually works. Demand specifics. Demand metrics. Demand outcomes. And treat the label with exactly the scepticism it deserves.
