How to safeguard user trust with AI without being a "speedbump"
Nobody wants to be their team’s AI “speedbump,” even if it’s better for the company and their users. Here’s what World Usability Day speakers recommended in this situation.
On Friday, I presented at UXPA Austin’s World Usability Day alongside Oen Michael Hammonds, Obianuju Okafor Ph.D., and Navita Deswal. The theme was trust: building AI processes and products that avoid user harm, work transparently, and do what they say they’re doing.
It’s a tall order for a technology that traditionally fails on all three counts–and made even harder in organizations where employees are expected to rapidly embrace AI.
So what is the ethical designer to do?
Across the talks, we shared three different strategies:
ETHICAL FRAMEWORKS: Agreeing on standards as a team for how AI will be adopted and implemented. We heard ideas in this vein from IBM, Microsoft, and Amazon, but these tend to be top-down approaches that may be hard for the individual designer to implement.
COMMUNICATION FRAMEWORKS: I personally advocate for “Disclosing & Differentiating” whenever communicating about AI tools and processes: “Disclose” exactly how you’ve used AI, and “Differentiate” between human and AI-aided processes. I favor this approach because it’s more bottom-up. You can control what you communicate about your own work and ask questions about others’–even without an agreed-upon, org-wide framework.
EXECUTIVE ALLYSHIP: Sometimes an executive can say what an individual contributor can’t. Executives can ask hard ethical questions and create the space for the additional work needed to build more responsibly with AI. So try to find these allies in your organization!
Sometimes, it is the designer’s role to introduce friction into the production process (shoutout to Jemma Ahmed for championing this idea). But it’s harder to do in a time when the value of design itself is being repeatedly questioned.
So keep these strategies in mind next time you find yourself needing to advocate for your users in an organization that’s rushing to implement AI a little too hastily.
Thanks again to Oen, Obianuju, Navita, as well as Aine E. and the rest of the UXPA Austin board for fostering this important conversation.
👉 If you’re trying to build greater appreciation for design’s role in AI adoption at your organization, consider booking me for a talk!