Every communication professional has experimented with ChatGPT by now. But most stop at ‘write me a press release’ or ‘summarise this text’. That’s like using a Formula 1 car to run errands. AI can do fundamentally more for communication work, if you know how to combine it with behavioural science.

That is not a theoretical promise. In the daily practice of communication teams, we see that AI only becomes truly valuable when it works alongside a clear understanding of human behaviour. The combination of Behavioural Design and AI, what we call BD×AI, opens up possibilities that most professionals don’t yet have on their radar.

Below you’ll find where those possibilities lie, why AI alone is not enough, and how you can start working smarter today.

Why AI alone is not enough

AI is exceptionally good at pattern recognition, processing large volumes of text and generating variants. But AI does not understand human behaviour. It doesn’t know why someone ignores an email, skips a call-to-action or abandons a form halfway through. AI produces output based on statistical probability, not on psychological insight.

The result? Without behavioural principles, AI generates generic, information-driven content. The texts sound good, but lack the psychological precision needed to actually change behaviour. You get a neatly formatted press release that nobody reads. A campaign message that informs but doesn’t activate. A newsletter that is correct but fails to prompt any action.[1]

The real breakthrough only comes when you feed AI the right mental models. When you know which behavioural mechanisms you want to activate, loss aversion, social proof, the framing effect, choice architecture, AI becomes an instrument that applies those mechanisms at scale. Not as a replacement for the thinking. As an accelerator of it.

Five AI applications that go beyond generating text

Most communication professionals use AI as a more efficient typewriter. That’s a waste. Here are five applications that show what is possible when you combine AI with a behavioural lens.

1. Audience analysis based on behavioural drivers

Traditional audience analysis revolves around demographics: age, education, region. But demographics barely predict behaviour. What does predict it are underlying motivations, fears, pain points and desires, the behavioural drivers behind decisions.

With AI you can analyse large volumes of qualitative data that were previously unmanageable: customer interviews, reviews, complaints, social media posts. Not to tally how often a word appears, but to uncover the underlying motivation patterns. Which fears keep recurring? Which expectations are being violated? Which trigger sets people into action?

A communication professional who masters this creates audience profiles that don’t describe who people are, but why they do what they do. That is a world of difference.

2. Testing framing variants

The framing effect is one of the most powerful principles in behavioural science: the same message leads to fundamentally different behaviour, depending on how you present it. ‘90% of customers are satisfied’ works differently from ‘only 1 in 10 is dissatisfied’, even though the factual information is identical.

AI enables you to generate ten, twenty or fifty framing variants for the same core message in minutes. Loss frames alongside gain frames. Social proof alongside authority. Emotional alongside rational. You can then test those variants, in A/B tests, in focus groups or through AI-driven analysis of expected effectiveness, and select the strongest frame in a data-driven way.[2]

3. Designing choice environments

Communication does not happen in a vacuum. Every message lands in a choice environment: a website, an app, a physical space, an email flow. And the design of that environment often has more influence on behaviour than the message itself.

AI can help you systematically map choice environments. By analysing a customer journey for friction points, by evaluating default options, by optimising the sequence of information presentation. Where do people drop off? Which default steers towards undesired behaviour? Where can a simple rearrangement make all the difference?

This is the domain of choice architecture, and AI makes it possible to analyse that architecture faster and more thoroughly than ever before.

4. Personalising behavioural interventions

Not everyone responds to the same behavioural mechanisms. One person is sensitive to social proof, another to loss aversion, yet another to authority. In the classic communication approach, you choose one strategy and hope it works for the majority.

With AI you can personalise by behavioural profile. Based on previously displayed behaviour, click patterns, response time, engagement, you can build models that predict which type of nudge is most effective for a specific recipient. One customer receives a loss frame, another social proof, yet another a scarcity principle. All automated, all evidence-based.

5. Campaign evaluation through a behavioural lens

Most campaign evaluations measure reach, clicks and conversion. Useful metrics, but they say nothing about why something works or doesn’t work. Was the message too abstract? Was social proof missing? Was there too much friction on the path to action?

AI can audit existing campaigns for the presence of proven behavioural principles. You input your campaign materials, landing pages, emails, ads, and AI analyses which principles are present, which are missing and where the greatest opportunities for improvement lie. No gut feeling. A systematic behavioural audit.

AI doesn’t replace the behavioural designer. It gives her superpowers. The combination of human insight and machine speed is where the magic lies.

The behavioural lens as AI superpower

There is a fundamental insight that connects all five applications: the quality of your AI output depends on the quality of your mental model.

If you ask ChatGPT for ‘a persuasive text’, you get a generic result. If you ask for ‘a text that activates loss aversion in decision-makers who display status quo bias, with a concrete CTA that minimises friction’, you get something fundamentally different. Not because AI has become smarter, but because you have steered it more intelligently.

This is the core principle of BD×AI: behavioural science makes your prompts better, your analysis deeper and your interventions more effective. You no longer speak the language of the marketer (‘make it catchy’), but the language of the behavioural scientist (‘activate social proof and reduce cognitive load’). And AI translates that with unprecedented speed into concrete output.

The behavioural lens is therefore not an optional extra. It is the multiplier that determines whether AI is a handy tool or a strategic weapon. Communication professionals who master this combination don’t just work faster. They work fundamentally smarter. Their campaigns are not only more efficiently produced, but also more effective at changing behaviour.[3]

Getting started with BD×AI

You don’t need to wait for a full implementation to experience the power of BD×AI. Here are three concrete steps you can take today.

  1. Audit your current campaign with a behavioural framework. Take your most recent campaign and assess it against five basic principles: Is there a clear default? Is there social proof? Is the friction path minimal? Is there a loss or gain frame? Is the timing aligned with a behavioural trigger? Use AI to accelerate this analysis, input your materials and explicitly request a behavioural audit.
  2. Rewrite one message in five frames. Choose a core message from your current communications and have AI generate five variants based on different behavioural principles: a loss frame, a social proof frame, a scarcity frame, an authority frame and an identity frame. Compare the variants and test which one resonates most strongly with your audience.
  3. Analyse customer feedback for behavioural drivers. Collect your last twenty customer reviews, complaints or interview excerpts. Feed them to AI with the instruction to search not for themes, but for underlying motivations, fears and barriers. The difference in output will surprise you, and you give yourself an audience profile that is actually actionable.

These three steps together take less than an hour, but they fundamentally change how you look at your communication work. You shift from broadcasting information to designing behaviour. That is precisely the shift that distinguishes communication professionals in 2026 and beyond from those who fall behind.

The tools are here. The science is here. The question is not whether AI is changing the communication profession, but whether you are the one who learns to make the combination with behavioural science. Read more about how we at SUE approach AI adoption through the lens of Behavioural Design.