Generative AU and the Future of UCD

Generative AI and the Future of UCD: Adapting to New Challenges

Discussions about generative AI seem endless—and while the topic may feel saturated, revisiting it in the context of user-centered design (UCD) and service delivery reveals critical opportunities and challenges worth exploring.

The Current Landscape of Generative AI

Generative AI is being increasingly evaluated for its potential to enhance research and public services. At the Ministry of Justice, for example, teams are exploring how generative AI can streamline user journeys, reduce duplication, and improve access to information—key pillars of effective service design.

While enthusiasm and investment in generative AI are high, the reality is more cautious. Most projects remain in the proof-of-concept phase, and feedback often reflects attitudes rather than real-world behaviors. Public trust in AI is low, and many people lack an understanding of how it works or how they might interact with it. In government and public services, unresolved questions about risk tolerance, error management, and human oversight signal that AI integration is still in its early stages.

Instead of declaring generative AI as the solution to user problems—or worrying about AI replacing jobs—it’s more productive to focus on adapting UCD practices to harness AI responsibly and effectively.

The Risk of ‘Solutionizing’ in UCD

Generative AI introduces a familiar challenge for UCD professionals: the risk of “solutionizing.” Many projects prioritize developing AI solutions, even before confirming they meet user needs. While experimentation is vital for exploring AI’s potential, there’s a danger in stakeholders prematurely assuming these proofs-of-concept validate AI as the ultimate solution.

This underscores the enduring importance of UCD in the “age of AI.” UCD professionals must ensure that user needs remain central, educating stakeholders not just about AI’s capabilities but also about why user-centered design leads to better outcomes.

To achieve this, UCD teams must prioritize ongoing user research and create opportunities for solution-agnostic ideation. Avoiding the “innovation trap”—assuming that the newest technologies inherently produce the best outcomes—requires openly acknowledging biases and finding creative ways to test assumptions. By doing so, decision-making becomes more transparent and adaptable, enabling cost-effective course corrections when needed.

How UCD Will Evolve

While the foundations of UCD will remain intact, generative AI will require adjustments to specific practices. For example, traditional usability testing might not fully address the variability of AI responses, which can differ even for identical user inputs. This unpredictability challenges conventional testing methods and demands new approaches.

User Research: Researchers may need to develop dynamic discussion guides that adapt to real-time interactions between users and AI.
Design Patterns: Designers may need to rethink linear user journeys, accommodating multiple, open-ended paths shaped by AI-driven interactions. New design patterns—or more flexible variations of existing ones—might be required to meet evolving user expectations.
Synthesis: Understanding AI behavior could become as critical as understanding user behavior. AI itself might be seen as a “user group” with its own role in the user experience.

Collaboration between UCD teams, data scientists, and AI developers will be essential. By working closely, these teams can better understand how generative AI interacts with users, ensuring its capabilities are leveraged effectively.

Rethinking Design Thinking

Generative AI might also shift how design thinking is applied within UCD. The traditional double diamond model emphasizes deep discovery and iterative solution exploration. However, when incorporating generative AI, it may be beneficial to experiment with AI’s capabilities earlier in the discovery phase, blending user problem exploration with rapid technical experimentation.

This approach would require guardrails to ensure user needs remain the priority, but it could lead to more innovative and practical solutions by aligning technical feasibility with user-centered insights from the outset.

Conclusion

Generative AI isn’t ready to replace jobs, but it does demand that UCD professionals evolve their practices. By adapting methods, increasing AI literacy, and holding innovation accountable to user needs, UCD teams can ensure that generative AI enhances, rather than detracts from, effective service design.

How do you see UCD adapting to the challenges and opportunities of generative AI? What other considerations should we anticipate? Let’s continue the conversation!

🔔🔔 Follow us on LinkedIn 🔔🔔