As AI systems take on greater decision‑making responsibility, the question of accountability becomes more urgent - and more complex. This fishbowl session explores what happens when AI‑driven insights go wrong: who owns the consequences, how risk should be governed, and what structures organisations need to ensure transparency, oversight, and responsible use. Drawing on emerging frameworks, evolving regulatory expectations, and real‑world implementation challenges, the discussion invites participants to examine how legal, compliance, and business leaders can safeguard decision quality when human judgment and machine intelligence collide.

Discussion points will likely include:

• Defining accountability in an AI-enabled organisation
  When AI-generated insights influence business decisions, where does accountability ultimately sit: with the user, the legal team, the business owner, the technology provider, or senior leadership? How should organisations establish clear ownership and governance before problems arise?

• Maintaining human oversight without losing AI's value
  How can organisations embed appropriate challenge, validation and human judgement into AI-assisted decision-making processes without creating so much oversight that the benefits of AI are lost?

• Managing AI risk before it becomes an incident
  What frameworks, controls and risk-management approaches are needed to identify issues such as bias, hallucinations, inaccurate recommendations, lack of explainability and regulatory non-compliance before they lead to poor decisions or reputational damage?

• Turning responsible AI governance into a competitive advantage
  As regulatory scrutiny increases and stakeholder expectations evolve, how can organisations move beyond compliance and build trust, transparency and accountability into their AI programmes in a way that enables innovation and accelerates adoption rather than restricting it?