Forget user experience. AI must focus on ‘citizen experience’

Tech giants and their AI-powered digital platforms and solutions can affect the destinies of world leaders, nation states, multinational corporations, global stock market, and individuals alike.

The creators of major digital platforms as well as the designers and developers of ubiquitous AI systems treat individuals as mere users, customers, or data points, oftentimes completely ignoring the individual’s role and rights as a citizen.

As a result, the individual users and customers are removed from the societal context with appalling consequences. The individual can unconsciously become a misinformation-spreading user. A misinformed customer can turn into a violent insurgent. Or she can be treated unfairly by a biased AI system while applying for a job or updating her insurance policy.

Why citizen experience design?

Now, as the societal impact of AI solutions is becoming obvious, their effects, as well as their design and development principles need to be considered from the point of view of citizens and society.

Already today, we’re seeing amazing work done in uncovering the effects of biased AI systems and their impact on various fields from healthcare to scientific research and from criminal justice to financial services. Simultaneously, we’ve been witnessing positive developments around data rights and practices.

But the regulation of tech companies, such as GDPR or data governance initiatives, isn’t enough. Similarly, the emerging field of algorithmic auditing doesn’t yet have sufficient means to directly affect AI development and its practices. Neither are the current AI ethics boards changing the course of AI development on a larger scale fast enough.

The most effective and sustainable impact on the field of AI will only be achieved by ensuring that the design and development of AI solutions is concretely guided by citizen-centric values and principles.

Previously, “citizen experience” has been seen as belonging solely to the field of public service. But no more. Today, we need a thoughtful citizen-centric approach that belongs to the general toolbox of every AI designer and developer.

Concretely, we need AI companies, data scientists, and designers that think and act citizen-first. We need citizen experience experts that bring the societal understanding into the core of product thinking, design, and development.

How to start thinking about citizen experience design for AI

So how can we create a basis for sustainable practice of citizen experience design that truly aims to create AI solutions that take into account the individual as a citizen, belonging to a wider fabric of society?

First, citizen experience design needs to be a multidisciplinary effort, bringing together social sciences, data science, and design. Data literacy and algorithm literacy are required for citizen experience design, i.e. concretely understanding the pros and cons of different data, and being able to assess the applications as well as potential effects of different algorithms. And this literacy can only be achieved by multidisciplinary approach.

Second, citizen experience design should help designers and developers to think of individuals as user-citizens and consumer-citizens. Citizen experience design should provide concrete tools for considering individuals as real people living in a real world, thus allowing a company to assess its product decisions in a wider societal context. Such tools would enable deeper user experience and customer experience design and data science practices.

Third, citizen experience design should affect all the elements of product design and development, from use cases and goal setting to applied metrics and user interface design, and from data pipelines and selection of AI technologies to user research and analytics. 

And fourth, the principles of citizen experience design must be created together with citizens. The co-creative practice will surface new insights that bring the citizen concretely into the center of things as an active force.

The founding principles of citizen experience design

It all starts with this: AI practitioners acknowledge the individual’s status as a full-fledged citizen and treat and respect her accordingly. AI solutions are never considered in the vacuum of a single product or platform.

Here are some concrete suggestions for further iteration:

AI systems should be designed and developed to guard the rights of the citizen. Algorithms are created in a responsible and transparent way. The AI system doesn’t treat citizens unfairly or endanger their immunity or integrity based on who they are.

A citizen’s data is handled and processed safely and responsibly. Personal data is not collected unnecessarily or used without explicit consent. Likewise, the citizen has to be made aware when she is interacting with, or being affected by, an AI system. An AI system should never try to fool or manipulate the citizen, for example by presenting itself as a human being or by optimizing a recommendation system for unhealthy addictive behavior.

When the citizen perspective is taken into account from the start, personal control of data must be thought of as a fundamental feature of any digital product. As the citizen’s rights to privacy are at the center of things, unnecessary AI-powered surveillance systems are out of the question.

AI systems should not promote behavior that violates any existing laws or the rights of other citizens. AI systems must respect the existing legislation and good manners. In short, AI designers and developers, or their AI solutions, do not decide independently what’s good, fair, or acceptable or what is lawful.

Citizen experience design empowers practitioners to proactively consider their solutions in the societal context through continuous dialogue with experts from different fields. AI solutions — recognized as socio-technological systems that are seamlessly intertwined with society — are continuously monitored, assessed, audited and iterated to mitigate potential problems or conflicts of interest early on.

AI systems should allow people to educate themselves about the use and effects of AI. When individuals are treated as full-fledged citizens, they also have to be held accountable for their use of AI solutions. For this, new citizenship skills are needed, including adequate data literacy, algorithmic literacy and digital media literacy. This requires effort both from citizens and AI practitioners. For example, the citizen could observe her data trails or exposure to algorithmic systems in an accessible manner.

Such educational transparency helps people to understand the motives and incentives of AI systems and their creators, building trust between citizens and AI developers. A citizen-centric and societally aware design informs citizens and empowers behavior and safety mechanisms that, for example, make harmful information operations easier to detect, mitigate, and even prevent.

A founding principle

Ideally, citizen experience design for AI should be a founding principle that concretely guides the design and development of AI solutions, not something that is used to assess or iterate the system retrospectively.

When looking at the big picture, it’s clear that citizen experience design will create new opportunities for AI innovations because the existing products as well as future solutions can’t ignore the individual’s multifaceted role as a citizen.

The core principles of citizen experience design must be created and iterated together. Let’s start today.

Jarno M. Koponen is Head of AI and Personalization at Finnish media house Yle. He creates smart human-centered products and personalized experiences by combining UX design and AI.