• AI for Good
• Bias & Ethics
• UX

Written by Amy Lynch, Senior Service Designer, Bank of England

Over the past few years, I’ve had many conversations with friends and acquaintances spanning a wide range of industries, including the civil service, agencies, consultancies, and the private sector.

Throughout these conversations, the topic of AI has come up time and time again, with work covering fascinating areas such as housing, planning, retail, food, and finance. I’m sure you have had these conversations too.

Each exchange has brought tales, hilarious anecdotes and occasional horror stories, of how different organisations are exploring AI’s potential.

Alarmingly though, so many initiatives seem to be a solution in search of a problem, rather than a solution meeting a need.

Unfortunately, we are quickly and acutely getting to know what poor AX (AI experience) feels like; whether it’s overzealous fraud detection, wacky voice recognition (ChatGPT still can’t grasp my accent), or incorrect outputs that we must manually, and exhaustingly, correct.

Think about humans first

While starting with the needs of humans isn’t the only way to deliver effective solutions, it is a tried and tested approach that can reduce risk, improve usability, and help organisations avoid some of the common pitfalls of AI implementation. Neglecting the user perspective tends to lead to challenges like:

• Workarounds and frustration: People will find the path of least friction, if it doesn’t work for them, they will start to bypass systems, processes and policies, reducing productivity potential and creating additional burden.
• Failure demand: If users can’t achieve their goals, they create additional pressure on support or other parts of the service, if a chatbot gives you silly answers, or responses you can’t trust, you will find another, more reliable channel, doubling up on support costs.
• Operational inefficiency: If Ai technologies are implemented without considering the whole experience, staff may need to step in to resolve unexpected issues, leading to increased manual interventions, eating into those anticipated efficiency savings.
• Costly redesigns: Solutions that don’t align with user needs often require significant rework later, whilst user centred design can seem like an unnecessary cost upfront, it often mitigates pain and complex unpicking later down the line.

There are tangible benefits of starting with the needs of humans:

Better journeys support higher adoption rates: when experiences align with user needs, people are more likely to embrace them. User dissatisfaction often stems from poorly planned service journeys, which can hinder the adoption of new products, services and systems.

Good usability supports productivity: helping users to be make less mistakes increases their productivity, research indicates that systems with high usability scores lead to substantially fewer user errors.

Increased trust: delivering reliable, user-friendly experiences builds confidence in the solutions, your team, or your organisation. Trust can be a huge influence, particularly in the use of solutions where Ai is a factor.

Compliance and accessibility: User-centricity can help ensure services meet legal and ethical standards. Making it easy for people to do the right thing often increases compliance with policies and governance.

How to identify human centred opportunities

An effective way of uncovering these opportunities is through exploring the user’s world.

Journey maps are a practical and structured way to do this. Advantageously many teams already have some form of these, or an alternative like service blueprints or process maps. When viewed through an Ai lens, these become powerful tools for spotting moments where technology could enhance, simplify, or even rethink parts of the experience.

While it’s tempting to jump straight to solutions, AI’s potential is vast and evolving too quickly to anchor thinking in specific tools. A more useful approach is to focus on AI’s fundamental capabilities;

• Generation
• Personalisation
• Automation
• Analysis
• Recognition
• Prediction

By framing AI in terms of what it can do rather than specific applications, it becomes easier to match the right capabilities to the right parts of the user journey. This shift in perspective helps move beyond novelty-driven experimentation and towards meaningful, user-centred innovation.

As AI continues to shape the future of services, its success will depend not just on technical advancements but on our ability to align it with genuine human needs.

A balanced approach

There’s no one-size-fits-all method for implementing AI. Some organisations might prioritise other factors, like cost savings or speed to market. But even in those cases, considering the user perspective can mitigate risks, reduce rework, and ensure a smoother experience for everyone involved.

Read More AI for good

Comments are closed.