Open any marketing tool’s homepage and you will struggle to find one that does not describe itself as AI-powered. The email platform that schedules your sends. The dashboard that shows you last month’s numbers. The chatbot that follows a script written by a human. All of it, suddenly, is artificial intelligence.
Some of it genuinely is. A great deal of it is software doing what software has done perfectly well for years, now wearing a more fashionable badge. There is even a name for the gap between the claim and the reality: AI washing. It is now widespread enough that regulators have started taking action.
For any business trying to spend its marketing budget wisely, telling the two apart matters. So here is a no-nonsense look at what counts as AI, what does not, and how to tell the difference without a computer science degree.
How AI washing crept into marketing
Part of the problem is that artificial intelligence is an enormous term. It stretches from the most basic rule-following software all the way to the systems behind tools like ChatGPT. When a word covers that much ground, it stops telling you anything useful about what you are actually buying.
The other part is incentive. Calling something AI sells. Back in 2019, the venture firm MMC Ventures reviewed 2,830 European companies classed as AI startups and found no evidence that AI was material to the business in roughly 40% of them. The same research noted that AI-labelled firms attracted considerably more funding than ordinary software companies.
That was before ChatGPT arrived and poured petrol on the fire. If the incentive to overclaim was strong then, it is far stronger now.
None of this means AI is fake or that the technology does not matter. It means the label has been stretched to breaking point, and a healthy dose of scepticism will serve you well.
What has been around for years and is now called AI
A surprising amount of what gets sold as AI is technology you may already use and understand. It is genuinely useful. It is just not necessarily artificial intelligence.
- Rule-based automation. If someone fills in a form, send this email. If a date passes, post that reminder. Useful, yes. But it is following instructions a person wrote, not thinking.
- Scheduling and workflows. Queuing posts and emails to go out at set times has been standard for well over a decade.
- Basic segmentation. Filtering your list by age, location or last purchase is a database query, not machine learning.
- A/B testing. Showing two versions and measuring which performs better is good practice and old news.
- Mail-merge personalisation. Dropping someone’s first name into a subject line is not, by itself, artificial intelligence.
- Decision-tree chatbots. If the bot only answers questions it was explicitly scripted to answer, it is a flowchart with a chat window.
There is nothing wrong with any of these tools. The problem only arises when they are priced and sold as something far more sophisticated than they are.
What is genuinely AI in marketing
Real AI in marketing can do more than simply follow a fixed set of instructions. Depending on the technology, it can generate new outputs, identify patterns, make predictions or adapt its responses based on data.
- Generative AI. The large language and image models that draft copy, summarise research or produce visuals. This is the technology most people now picture when they hear the term.
- Predictive machine learning. Models that learn from data to forecast which supporters might lapse, which leads are worth pursuing or what a customer is likely to do next.
- Recommendation engines. Systems that work out what to suggest based on patterns across large numbers of behaviours, rather than a simple rule you wrote yourself.
- Natural language processing. Tools that can analyse open-text feedback and identify sentiment, themes or intent.
- Real-time optimisation. Systems that adjust decisions such as bids, targeting or recommendations based on incoming data.
The important distinction is not whether a tool uses the letters “AI” in its marketing. It is what the technology actually does and whether that capability is genuinely different from standard automation.
Four questions that cut through the marketing
You do not need to be technical to test a bold AI claim. You just need to ask the right questions and listen for a straight answer.
- Does it learn, predict or generate outputs from data, or does it simply follow fixed rules someone set up?
- What can it do that a well-built spreadsheet or a standard automation could not?
- What data does it rely on, and what happens if that data is limited or unavailable?
- Can you explain, in plain English, what the technology actually does?
If the honest answer comes back as “it follows the rules we configured,” you are looking at automation. That is completely fine. Just do not pay AI prices for it.
Why this is more than pedantry
This matters for three practical reasons.
The first is money. Rebadged basics can carry a premium simply because the AI label justifies a higher price. Knowing what you are buying helps keep your budget where it belongs.
The second is disappointment. When a tool is sold as intelligent and then behaves like a rigid script, the gap between promise and reality erodes trust quickly, both in the tool and in whoever recommended it.
The third, increasingly, is legal risk. Regulators have noticed. In September 2024, the US Federal Trade Commission launched Operation AI Comply and has taken action against companies accused of making deceptive claims about AI capabilities. In one widely reported example, an app marketed as AI-powered checkout was alleged to rely heavily on human workers processing orders.
As the FTC put it, “there is no AI exemption from the laws on the books”. UK advertising and consumer protection rules follow the same basic principle: if you make a claim about what your technology can do, you need to be able to support it.
Our take
We are not anti-AI. Used well, it is genuinely valuable, and we help clients put it to work where it earns its place. What we are against is hype dressed up as substance, because it wastes money and chips away at trust, and trust is the thing values-led businesses cannot afford to lose.
So when a tool tells you it is powered by AI, it is worth pausing and asking a few honest questions. Some of it is the real thing and well worth having. Some of it is good software with a fashionable label.
Knowing which is which is simply how you spend wisely.
Want a straight answer about your marketing technology?
If you are weighing up tools that promise the world and want an honest view of what is genuinely worth your budget, we would welcome a conversation. At NAMA, we help values-led businesses and organisations cut through the noise and invest in marketing that actually works.
