
Too Much Autonomy! — When to Bring in the Human to Keep Agentic AI Accountable?
Interactive workshop (13 Feb 2026) exploring when to bring humans into agentic AI systems to maintain accountability, safety and control. Learn why autonomous agents behave unpredictably, how a human-in-the-loop stops ‘rogue’ behaviours, and practical, human-centred coding practices to keep AI reliable. Through live demos with everyday robots and computers, attendees will step in as the crucial human, test interventions, and see hands-on examples of preventing privacy breaches and failure modes. Ideal for students, researchers, AI builders, educators and curious members of the public interested in AI, automation, agents and safe design.