Inteligência Artificial e IA Generativa
Oradores que descodificam o impacto real da inteligência artificial em indústrias, força de trabalho e vantagem competitiva
Speakers Associates represents 355 speakers on Inteligência Artificial e IA Generativa, including Kemal Apaydin, Olivier Sibony, Rahaf Harfoush, Purna Virji, Itai Green, Limor Ziv, Tom Goodwin, Daniel Trabucchi & Tommaso Buganza, Katja Schipperheijn e Diana Verde Nieto.
Employees are arriving at work already exhausted by their relationship with technology, then asked to absorb AI on top of it. Attention is fragmented, identity is leaking into datasets, and the human costs of always-on connection are showing up in engagement scores and mental health budgets. Leaders are running wellbeing programmes that do not touch the actual mechanism causing the harm.
Most mid-sized European companies have run AI pilots. Few have moved them into operating reality. Boards are stuck between vendor pitches, internal scepticism, and a workforce already split between people who use AI daily and people who don’t.
Marketing budgets are getting bigger while the proof that any of it works is getting weaker. Viewability metrics inherited from a decade ago tell buyers an ad was technically on screen; they say nothing about whether a human noticed it. The gap between paid impressions and commercial outcome is now the single largest unmanaged risk on the marketing P&L.
Most breaches do not start with a flaw in the firewall. They start with a person who answered the wrong email, trusted the wrong voice, or approved the wrong wire. Security spend keeps rising while the attacker keeps targeting the human layer, and most organisations still treat that layer as a training problem rather than a behavioural one.
Most boards have approved an AI strategy. Far fewer can explain how their models make decisions, where the bias sits, or what they will say to a regulator when one of those decisions is challenged. The gap between procurement and accountability is widening, and the answer is not another tooling vendor.
Most organisations have already run AI pilots. The harder question is what happens after the proof of concept ends. Procurement standards stay unclear, accountability for AI-assisted decisions is unassigned, and the governance frameworks people quote in slides do not survive contact with real workflows. Leadership teams cannot say with confidence which decisions AI should be trusted with and which it should not.
Every senior leader has been told that technology ethics matters. Very few have been given a way to make ethics decisions that also survive a board review or a regulator’s letter. In AI, surveillance, biometrics and the platforms now embedded in every function of the business, the question is no longer whether to worry about ethics, it is how to make defensible choices at the speed the technology is moving, with the operating, legal and reputational consequences those choices carry.
Most senior teams have run their first generative AI pilots and stalled. The technology is general-purpose, but the operating decisions are not: which workflows to redesign, which tools to standardise on, where hallucination is tolerable and where it is not. The question is no longer whether to adopt, but how to convert curiosity into measurable operating advantage without ceding judgement to the model.
Artificial intelligence is built to imitate the brain, yet most leaders backing it cannot say how the brain actually works. The only proven model of general intelligence is still biological. Understanding how it remembers and finds its way is becoming useful in judging what machines can and cannot do.
Boards are now accountable for AI decisions they do not fully understand. Regulators, customers, and employees expect defensible governance, but most companies still treat ethics as a slide at the end of the deck. The gap between AI ambition and AI accountability is where reputational, legal, and operational risk now compounds fastest.
Financial firms are under pressure to put generative and agentic AI into regulated work without breaching rules, losing trust, or building tools advisers ignore. Most boards can describe the opportunity; far fewer can describe the operating model, the controls, or where an agent stops helping and becomes a liability. The gap between AI ambition and deployment that creates value without eroding the business model is where most programmes stall.
Most boards are now asked to approve AI decisions they do not understand, under regulation that is still settling. The hard work is no longer pilots. It is deciding where AI belongs in the operating model, who is accountable when it fails, and how to defend those choices to regulators, customers and employees.