How Belgium’s financial sector can lead the AI journey

On Thursday, 26 June 2025, KPMG in Belgium welcomed a diverse group of professionals and thought leaders to its Zaventem offices for an engaging half-day event focused on a topic that is rapidly transforming the financial landscape: Trusted Artificial Intelligence (AI).

The event brought together a distinguished panel of experts representing the financial, legal, academic, and entrepreneurial sectors. Participants explored the future of AI in banking and insurance, with a particular emphasis on governance, trust, regulatory alignment, and operationalization of AI in real-world settings. The backdrop of the EU AI Act, combined with emerging technologies like generative AI and autonomous agents, made this conversation both timely and urgent.

Bart Becks, our guest speaker, delivered a bold vision for Europe’s tech future, emphasizing the need to accelerate innovation through targeted investments and a streamlined regulatory framework in key sectors such as AI, deeptech, and defense.

He highlighted the importance of building a truly integrated European innovation ecosystem, where corporates, policymakers, venture capital, academia, and entrepreneurs work together inspired by success stories from Estonia and the Nordics.

Bart also shared concrete and coordinated EU actions planned by 2026, such as the 28th regime, the European Business Wallet, and the Blue Carpet Initiative - all aiming to turn Europe into fertile ground for startups and scaleups.

Bart Van Rompaye, Head of Artificial Intelligence at KPMG, set the tone by framing the delicate balance at the heart of “trusted AI”: how to maximize value, ensure regulatory compliance, and mitigate risk all while earning and maintaining stakeholder trust. His presentation unpacked the building blocks of a robust AI governance framework and called for a more human-scale approach that integrates governance intuitively into the systems people already use.

The panel discussion, moderated by Heleen De Jager, featured experts from across the financial and AI spectrum:

• Françoise Guebs (National Bank of Belgium) outlined the regulator’s perspective on AI risk and trust, highlighting the importance of explainability and the need for early alignment between AI systems and regulatory requirements.
• Laurence Van Meerhaeghe (Febelfin) emphasized the compliance burden faced by smaller financial institutions and advocated for clearer, more streamlined regulatory support to avoid “regulatory fatigue.”
• Hannes Heylen (Belfius Insurance) offered a successful use case in fraud detection, where enhancing the interpretability of AI outputs significantly improved business adoption and generated measurable financial benefits.
• Dimitri Devlamminck (Indosuez) shared lessons from implementing AI governance across multiple jurisdictions, stressing the importance of group-level AI policies and centralized expert boards to streamline assessments.
• Prof. David Martens (University of Antwerp) cautioned against underestimating the emerging risks of agentic AI, including reduced human oversight and accountability, especially as AI systems begin to act autonomously.
• Bart Becks (Serial entrepreneur and keynote speaker) assessed Belgium’s positioning on the compliance–innovation spectrum, encouraging leaders to build clarity around risk appetite and foster a culture where governance and innovation reinforce - rather than contradict - one another.

Several key messages resonated throughout the morning:

1. AI must be trustworthy by design. 
   Embedding trust into AI is not a check-the-box exercise. It requires strategic alignment between the organization’s values, business goals, and technical implementations. Trustworthy AI is not just compliant ,it is also transparent, explainable, fair, and accountable.
2. AI governance needs to evolve.
   Traditional governance models often fail to keep up with the pace of AI innovation. Instead of rigid frameworks, organizations should invest in agile, human-centric governance systems that are embedded into daily operations and supported by strong cross-functional collaboration.
3. Explainability is a game-changer.
   In practical use cases like fraud detection, the addition of explainability tools to support non-technical stakeholders proved to be pivotal in achieving user buy-in. When business teams understand how AI works, trust increases and so does adoption.
4. AI regulation is in flux — act proactively.
   With the EU AI Act entering into force soon, financial institutions must not wait. The landscape will continue to evolve, and those that build scalable, future-proof AI governance now will be in the best position to adapt.
5. AI agents are the next frontier.
   As AI systems move from reactive tools to autonomous agents, they introduce new complexities around accountability, data privacy, consumer risk, and systemic bias. Preparing for this shift requires updated risk assessments, new audit techniques, and deeper technical literacy across the organization.
6. Trust starts with culture.
   From leadership to development teams, building a culture of trust is essential. Boards must clearly define their risk appetite, promote transparency, and create environments where governance is seen as an enabler, not a bottleneck.

As AI becomes more embedded in core financial operations, the stakes will only rise. Institutions must take deliberate action today to lay the foundations for trustworthy, scalable, and future-ready AI systems.

At KPMG, we’re here to support you in that journey, whether by assessing your AI maturity, building governance frameworks, preparing for regulatory changes, or strengthening security and ethical oversight.

Contact our AI & Financial Services team to explore how we can help you shape a resilient and responsible AI strategy.