In 2024-2026 every roadmap seems to require "something with AI". Investors, boards and competitors push for visible features — chatbots, recommendations, automatic generation. The designer often finds themselves designing interfaces for functionality whose utility or safety has not been assessed.
AI washing is the most immediate risk: labelling as "intelligent" what is an if-then rule, improved autocomplete or a cosmetic API integration. From a UX perspective, this erodes trust: the user expects capabilities the product does not have, and when disappointed, blame falls on the experience — not marketing.
Tech FOMO pushes teams to skip fundamental steps: testing with real users, edge case evaluation, analysis of who gets excluded. "Shipping AI" becomes an end in itself. The designer can counter with concrete questions: what problem does it solve? What happens when it gets it wrong? Who can challenge the output?
The designer's professional responsibility goes beyond delivery. Codes of ethics such as the World Design Organization's emphasise that design must serve the wellbeing of people and society. Integrating AI without assessing its impact is not neutrality — it is a choice.
It is not about opposing every innovation. It is about asking for evidence, proposing alternatives (including not using AI), documenting trade-offs and bringing concrete failure examples into meetings — racist filters, chatbots giving dangerous medical advice, onboarding that excludes neurodivergent users.
Being "the ethical spoilsport" is exhausting. But the designer who understands implications of bias, transparency and consent has negotiating leverage: they can translate abstract risks into concrete UX scenarios that product managers and stakeholders understand — and sometimes fear.