Future AI

Chairpersons welcome and opening remarks

As AI regulation accelerates, large organisations must go beyond checklists to build scalable, cross-functional governance. This session explores how leading companies are aligning legal, risk, and technical teams to ensure compliance, oversight, and responsible innovation across the AI lifecycle.

• Key updates on the EU AI Act and anticipated U.S. regulatory shifts

• How multinationals are adapting governance structures to stay compliant

• Building alignment between legal, risk, compliance, and technical teams

• Creating scalable frameworks for AI oversight and auditability

• Best practices for embedding responsible AI into product development lifecycles

• Balancing innovation with control in large-scale AI programmes

As AI moves from experimentation to large-scale deployment, responsible governance is no longer optional, it is strategic. From compliance with the EU AI Act to evaluating financial risk, auditability, and long-term ROI, organizations must think ahead of implementation, not after disruption. This session explores how leaders assess emerging AI technologies through the lenses of regulation, accountability, and scalability. We will also examine the future of work, where humans and AI must operate in partnership, and why responsible AI will define competitive advantage in the years ahead.

Many organizations have data but still struggle to turn it into better decisions. In this session, Tamires Beltrão shares how data insights can move beyond reporting to become a true driver of performance and strategy.

Drawing on more than a decade of experience across Latin America and Europe, and her current role leading Data Insights at the Port of Rotterdam, Tamires explores how to connect analytics, KPIs, and advanced methods to real business questions. She reflects on when sophisticated techniques like AI and machine learning add value and highlight why relevance, clarity, and strong collaboration with the business are the keys to making data stick.

A practical discussion on how enterprises are building out their AI assurance stack – covering risk classification, third-party tooling, and internal controls for ongoing monitoring, logging, and auditing.

What does a modern AI risk management toolkit look like?
Buy vs build: approaches to governance and compliance tooling
Embedding assurance into DevOps and model monitoring pipelines

What 150+ enterprise deployments and hundreds of millions of customer conversations taught us about getting AI agents from pilot to production

AI models have become dramatically more capable, yet most enterprise agents still never make it past the pilot, or get pulled once they’re live. Recent research shows that 74% of enterprises have already rolled back a live AI customer-service agent. Benchmark accuracy keeps climbing, but reliability has barely moved. The gap isn’t intelligence. Organisations treat agents as features to ship, when they are systems to operate.

Drawing on hundreds of millions of agent conversations across 150+ enterprise deployments, Csaba Tamas, CPO of Parloa, shares what actually breaks when AI agents meet real customers, and what the teams that succeed do differently. He explains why the most mature organisations roll back their agents more often, not less. He shows why courts in Canada and Germany now hold companies liable for every word their chatbots say. The session closes with a look at who will build the next generation of enterprise agents, and at the new customer on the line that most companies aren’t ready for.

Key takeaways:
– Why agents that look perfect in a demo fail in production, and the maths of why reliability drops as tasks get longer

– Why containment rate is the wrong KPI, and how to measure the value an agent creates rather than the cost it avoids

– How to make agents resilient: guardrails, testing every change, and a fix loop measured in hours, not months

– What actually raises automation rates: backend actions, knowledge coverage and improvement cycle time

– Why the people who own customer outcomes, not engineers, must be able to change the agent

– How to run AI agents like a workforce: scoped, onboarded, supervised and coached

As organizations race toward autonomous, AI-driven operating models, C-suite leaders face critical decisions on strategy, governance, talent, and technology. This panel brings together enterprise executives who are at the forefront of building agentic organizations, where AI agents collaborate with human teams to drive productivity, innovation, and real-time decision-making.

Panellists will explore frameworks for scaling autonomous agents responsibly, aligning investments with business outcomes, navigating emerging security and compliance challenges, and reimagining workflows and culture. Attendees will gain a practical playbook for leading their organizations through the next phase of AI transformation and capturing sustainable competitive advantage.

The talk will detail how Agentic AI is deployed in weeks instead of months/years by combining the power of the Wonderful platform with last-mile FDE delivery. We will shortly go over the architecture of our Agentic orchestration platform that connects to any intelligence provider, can be used self-serve via the UI or headless with your favourite coding tool and that can be hosted multi-cloud or even on-prem. All delivered by the best engineers we can find in the Dutch market to deploy agentic solutions shoulder to shoulder.

Transparency is no longer a ‘nice to have’-it’s a strategic differentiator.

Learn how leaders in finance, healthcare, and logistics are using explainability and communication to earn trust from users, partners, and regulators.

Key Discussion Points:
– What transparency looks like in real-world enterprise settings

– How to communicate model behaviour across internal and external stakeholders

– Examples of organisations using openness to win customer and investor confidence