AI-etik och ansvarsfull teknik
Talare som granskar de mänskliga konsekvenserna av algoritmiska beslut, dataetikett och framväxande teknik
Speakers Associates represents 149 speakers on AI-etik och ansvarsfull teknik, including Kemal Apaydin, Rahaf Harfoush, Limor Ziv, Harriet Farlow, Saakshar Duggal, Dr Sidney Shapiro, Tina Stowell, Timandra Harkness, Dame Wendy Hall och Susi O’Neill.
Boards have approved AI investment. Most have not yet decided what good looks like. The question is no longer whether to deploy AI, but how to deploy it without inheriting failure modes that legal, regulatory and reputational teams cannot defend later.
Most organisations deploying AI have optimised for capability, not accountability. Algorithms now shape hiring, lending, clinical diagnosis, and criminal justice at scale – but the governance structures to challenge them barely exist. The gap between what a model optimises for and what an organisation is actually accountable for is where the real risk lives.
AI systems are going into production faster than anyone is checking whether they can be attacked. A model can be poisoned during training or manipulated after deployment, and most security functions have no test for either. That risk sits between the data science team and the CISO, and neither owns it.
Most inclusion programmes never reach the decisions that matter. Hiring, promotion and performance calls keep producing the same outcomes, and those decisions are increasingly made by AI. The hard question for a leadership team is what to do differently in the next cycle, and what to check before trusting the algorithm that runs it.
Most security programmes are built by defenders who have never run an intrusion end to end. The result is a set of controls that look complete on a slide and fail in the specific places an experienced attacker already knows how to find. Closing that gap requires an honest account of how hacker groups form, choose targets, and move through a network, told by someone who did it.
Regulated institutions know how to pass a compliance review. The harder test is whether their governance could catch an ethical failure before it becomes a reputational one. A diversity policy and a structurally inclusive institution are not the same thing, and the distance between them is now being measured.
Boards now make capital and operating decisions inside a system where geoeconomic competition, supply shocks, technological disruption, and political fracture move faster than the institutions designed to manage them. Most leadership teams understand each risk in isolation. The harder problem is reading how they compound across regions and sectors, and what that means for growth, capital allocation, and the next decade.
Boards know AI is not optional. What they do not know is which of the dozen initiatives on the deck will compound into advantage, and which will sink six quarters of budget into pilots that never scale. The gap is not ambition, it is a repeatable way to decide where the organisation actually stands and what to do next.
Most boards now treat AI as a strategic line item, but few know how to translate it into operating advantage without tripping the regulators, the workforce, or the customer. The gap between AI ambition and AI deployment is widening, not closing. Leaders need someone who has sat on both sides: the commercial side that has to ship, and the governance side that decides what shipping looks like.
Most leadership teams have an AI strategy. Far fewer have changed how the business runs. The gap between stated intent and operating-model impact is where executive teams stall, and where the investment case quietly unravels.
Boards are being asked to approve AI strategies they cannot evaluate. The architects of frontier systems openly say they do not fully understand what their models can do, yet executives are expected to deploy, govern and disclose around them. The shortfall is not technical literacy. It is a working theory of where the technology is heading and what that means for capital, headcount and liability.
Most organisations have run AI pilots. Few have moved from pilot to operating capability. The gap is rarely the technology; it is the absence of a structure that connects model choice, team design, ethics, and day-to-day decision rights across the business.