How AI will influence the Job ob Accessibility Proffessionals

Today we are addressing the question of how artificial intelligence (AI) will change the role of accessibility experts.

AI as a Tool for Routine Tasks

The capabilities of AI are increasing rapidly, especially in the fields of design and programming. In just a few years, AI will take over many tasks that we still perform manually today. These include, in particular:

Testing components and applications.
Generating optimization suggestions.
Retesting resolved issues.

These traditional tasks can be effectively handled through pattern recognition and statistical methods. In the future, AI will directly support tools like Figma or developer tools in generating accessible code and ensuring accessibility.

To put this into perspective: we are not there yet. Many providers of so-called “accessibility overlays” promise magical AI-driven solutions, but upon closer inspection, much of this turns out to be marketing. In reputable tools such as Accessibility Cloud or PDF Accessibility Checker, there are already useful AI features, but they are far from replacing human work. As long as tools remain at their current level, there is no need for concern about job security. It is also questionable whether AI will ever be able to create a fully accessible complex application entirely without human assistance.

Nevertheless, one thing is clear: we do not know how good these tools will be in 1–2 years. A significant leap in quality over the coming years is quite possible—and even likely.

The Shift: From Testing to Consulting

The profession will shift significantly: away from pure testing and toward strategic consulting. Experts will increasingly be needed where AI reaches its limits:

Decision-making in edge cases: AI often cannot make clear judgments. Experience is needed to assess the impact on assistive technologies (such as screen readers).
Context analysis: Making a single component accessible is feasible for AI. The real challenge lies in how all elements interact. An expert must ensure correct focus behavior and meaningful error messages.
Governance and strategy: In large organizations, the focus is not just on a single app but on entire software ecosystems. Guidelines must be established to ensure consistency. Accessibility needs to be firmly integrated into processes and frameworks.

New Requirements for Experts

This shift makes the field more complex and more exciting. In addition to expertise in accessibility, new skills are becoming increasingly important:

Communication and presentation: Experts must explain requirements clearly and understandably.
Empathy and persuasion: Stakeholders from legal, design, and management must be engaged and convinced of the importance of accessibility.
Systems thinking: Collaboration with UX, design, development, and compliance teams becomes central.

In summary: AI will take over routine tasks, but strategic direction and final responsibility will remain human responsibilities.

Quality Assurance for AI-Generated Content

A central issue with today’s AI lies in how it works: it is based on probabilities. This means that results are not consistently stable. It can happen that the first version of a piece of code is accessible, while a second version—after minor changes—reintroduces barriers. Incremental degradation is a well-known phenomenon in current AI tools.

Accessibility experts must therefore ensure that the AI tools themselves operate reliably. It is not enough to purchase a license and hope for “magical” results. For an AI to adhere to corporate design or accessibility standards, it requires a solid data foundation—often referred to as a knowledge base or as part of a Retrieval-Augmented Generation (RAG) approach. This knowledge base contains the specific instructions and component libraries that the AI must follow.

In the future, experts will need to engage deeply with prompting technologies. They will guide AI through precise instructions and continuously refine the knowledge base. Instead of manually checking everything, the focus will shift to analyzing AI output:

Where did the AI make mistakes?
Was the prompt or the knowledge base ambiguous?
Are there conflicting requirements within the knowledge base?

The goal is to optimize the automatic generation of accessible code so that the need for manual corrections is reduced.

User Experience Instead of Compliance

If AI frees up capacity previously spent on manual testing, we can finally focus on the most important topic: usability.

An application can be formally accessible and still be unusable. Many applications today are simply too complex—whether for people with disabilities, older adults, or less tech-savvy users. What we produce today is often inadequate in real-world use.

This is where the future lies:

Usability heuristics: We must evaluate whether the user experience is truly intuitive.
User testing: We need real testing with people with disabilities and other vulnerable groups.
Collaboration with UX research: Accessibility experts should work closely with UX colleagues to design inclusive testing methods.

By focusing less on technical checklist testing and more on actual usability, we can significantly improve the quality of digital products.

Long-Term Strategy Instead of Short-Term Projects

A major issue for the future has little to do with technology itself: the development of accessibility policies. The core problem today is that accessibility efforts often restart from scratch. Experts are frequently brought in only for a relaunch. Once the project is completed, their involvement ends. In the following years, accessibility typically deteriorates gradually—until the next relaunch.

Given the vast number of applications that public authorities and companies operate today, this fragmented approach is no longer sustainable. We need continuous governance. Accessibility experts must no longer be temporary project contributors. They need to be integrated into organizational structures—either as employees or as strategic consultants.

Presence matters: Simply by being consistently “at the table,” experts ensure that accessibility receives the attention it deserves. Accessibility must become part of corporate governance. This means embedding it into all processes rather than isolating it in silos. Without such overarching control, product quality will decline over time rather than improve—as is evident in many inadequate applications on the market today.

Conclusion

The role of experts is undergoing a radical transformation: from purely technical testers to strategic designers and communicators. We use AI as a tool for routine tasks, freeing up mental capacity for what truly matters: usability, governance, and genuine inclusion.

The advantage is that a broader range of skills will be needed: both technical skills (such as AI prompting and quality assurance) and communication and organizational skills associated with management. This means that people with different strengths and abilities have equal opportunities to contribute meaningful and important work. However, it is crucial to start preparing for these responsibilities now.

Our task is not to demonize AI, but to stay ahead of the curve and actively guide it. If we influence AI and steer its development in the right direction, we have a strong chance of making technologies significantly more accessible than they are today.