Step Aside, Human, But Do Not Misstep: How to Use and Govern Agentic AI Safely

Little helpers are emerging everywhere: they can deal with employee queries, provide you with a draft email with reference to indicated resources, and even try to find compatibilities between you and your estranged other half. Our panel will explore which parts of their lives humans are delegating to AI agents. We will share practical examples of the opportunities that agentic AI open and gaze into the crystal ball to predict how they may be taken up. We will discuss if the risks that AI may pose are amplified in the case of agentic AI, as AI agents’ autonomy to take actions impacts the physical and digital world. We will touch on the challenges that AI agents present across industries (e.g. reproducibility and compliance). Against the background of topical unacceptable and high risk uses, we will offer pragmatic steps that the audience can implement to benefit from agentic AI and reduce the risks – e.g. by augmenting human intelligence and using adversarial human-AI collaboration. We will offer solutions for ethical AI governance in compliance with the myriad of digital and data laws impacting AI: old and new, and not necessarily consistent. 

What you will learn:

• Explore the current state of play of agentic AI, including specific opportunities, risks and challenges.
• Share practical steps that developers and deployers can undertake to govern and use AI agents safely.
• Understand the cross-section of various digital and data laws that apply to agentic AI to be ahead in compliance efforts.