Éthique de l'IA et technologie responsable
Des conférenciers qui interrogent les conséquences humaines des décisions algorithmiques, l’éthique des données et les technologies émergentes
Speakers Associates represents 149 speakers on Éthique de l'IA et technologie responsable, including Kemal Apaydin, Rahaf Harfoush, Limor Ziv, Harriet Farlow, Saakshar Duggal, Dr Sidney Shapiro, Tina Stowell, Timandra Harkness, Dame Wendy Hall, et Susi O’Neill.
Most organisations are spending heavily on AI and still producing the same ideas they produced last year. The bottleneck is not the model or the tooling; it is the quality of human judgement brought to the work. The question senior leaders keep returning to is how to get original thinking and technological leverage from the same teams at the same time.
Boards are now expected to have a view on AI, online manipulation and digital trust without having lived inside any of those worlds. The gap between what executives understand about the internet and what is actually happening on it has become a governance problem, not a technology problem. Most strategy documents treat that gap as a training issue. It is closer to a credibility issue.
Most leadership teams have an AI strategy that describes adoption. They do not have one that describes consequences. The systems being deployed across defence, finance, and healthcare are no longer tools that can be audited line by line, and the gap between what an executive can authorise and what the underlying technology actually does is widening month by month.
Boards are being asked to make consequential bets on generative AI without a stable read on what the technology can actually do, what it cannot, and what its deployment will mean for the workforce. Most executive briefings collapse into either hype or alarm. Leaders need a sober technical interpreter who can separate marketing from mechanism, and tell them which decisions matter now.
Boards are signing off on AI deployments faster than their organisations can govern them. Privacy, consent, and data lineage have moved from compliance topics to live commercial risks tied to model training, customer trust, and regulatory exposure. Most leadership teams have no shared language for deciding which uses of data are defensible and which are not.
Trust between brands and the people they sell to has eroded faster than marketing functions can rebuild it. Generative AI now writes the copy, targets the audience and shapes the campaign, and consumers know it. The commercial question is no longer how to be seen, but how to be believed.
Every organisation now has a digital transformation strategy. Very few have the executive fluency to decide which emerging technologies actually deserve investment, which are years away from being usable, and which belong on the regulator’s desk rather than the roadmap. The cost of getting that distinction wrong, in smart-city programmes, public-sector IT and corporate digital strategy, is quietly absorbed as failed projects and stranded spend.
Artificial intelligence is moving from pilot to protocol inside hospitals, space agencies, and infrastructure programmes, and most leadership teams are still arguing about what is real and what is theatre. The cost of getting this wrong is not slower innovation. It is patient harm, missed regulation, and capital deployed against the wrong assumptions. Boards want a translator who has actually built and deployed clinical AI, not a commentator describing it from the outside.
Boards have signed off on AI ambitions that the operating business has no idea how to execute. Pilots multiply, vendor decks pile up, and the gap between strategy slides and what customers actually experience keeps widening. The job leaders need help with is choosing where AI changes the commercial model, and where it is noise.
Autonomous AI agents have started acting on people’s behalf, sending messages, booking, buying, and making decisions without a human in the loop. Organisations now have to decide how much of that autonomy to hand over, and who answers for it when an agent gets something wrong. The technology is moving faster than the controls and oversight meant to govern it.
Boards are being asked to make calls on artificial intelligence and health technology before the evidence base has settled. Most senior teams have a strong grasp of the hype cycle and a weak grasp of what the science actually supports, where the ethical exposure sits, and which innovations will reach customers and workforces inside the planning horizon. The gap between confident vendor pitches and defensible internal judgement is widening.
Most organisations evaluating AI can assess technical performance. Few can assess what AI systems do to decision-making structures and accountability lines once deployed. That gap, between what AI promises and what it changes about how organisations operate, is where governance risk accumulates before it becomes visible.