Registration & Welcome Coffee

Opening speech

Moderator: Yugo Neumorni, CIO Council President

Title Session:

Synopsys:

Speaker: Igor Babitsch, ServiceNow

Title Session:

Synopsys: 

Speaker: Dell Technologies rep

PANEL1: The AI Reckoning: From Pilots to P&L

$2 trillion spent, most pilots showing zero return — is enterprise AI a bubble, or just badly deployed?

Synopsis

The hype cycle has met the balance sheet. Widely cited research in 2025–2026 found that roughly 95% of enterprise generative-AI pilots delivered no measurable financial return, even as global AI spending approached $2 trillion — and analysts expect a large share of AI projects launched in 2024–2025 to be defunded for failing to prove value. Yet a minority of organisations report triple-digit ROI. So what separates the two? This panel cuts through the noise to ask what real, booked value from AI looks like, why so many initiatives stall between pilot and production, and how leaders should budget, measure and communicate AI in a year when patience — and capital — is running out.

Key discussion threads

• The "95% zero-ROI" studies — myth, measurement problem, or genuine warning?
• How should a CIO calculate the real ROI of AI?
• Which AI use cases have actually delivered measurable results?
• Why pilots die between the lab and production: legacy integration, data readiness, missing business owners.
• Defining and measuring return on intelligence: productivity, revenue, risk, cost-to-serve.
• The build-vs-buy-vs-wait dilemma and the discipline of killing failing projects early.
• Bubble economics: what happens to budgets, vendors and expectations if the correction comes.

Panelists: 

• Gabriela Stănică, Chief Information, Data Officer and E-commerce Director, Carrefour Romania, CIO Council Member
• Oana Laza, Services & Solutions Delivery Director – Marketing, Insights and R&D, Imperial Brands PLC, CIO Council member
• TBC
• TBC
• TBC

Moderator: Andreea Ionita, Digital & Innovation Director CEE, CRH, CIO Council Member

Coffee Break 

Title Session:

Synopsys:

Speaker: Yossi Altevet, DeepKeep, CTO and Co-founder

Title Session:

Synopsys:

Speaker: TBC

PANEL2: The Agentic Enterprise: When Software Starts to Act

AI agents are moving from demo to production — but when an autonomous agent makes the wrong call, who is accountable?

Synopsis

2026 is the year the industry shifts from chatbots that answer to agents that act — booking, buying, reconciling, coding and orchestrating other agents. Adoption surveys report a majority of enterprises now running at least one agent in production, with open protocols emerging as the standard for agent-to-agent communication. But autonomy raises the stakes: multi-agent systems are harder to observe, test and control, and a widely noted "governance gap" means capability is racing ahead of oversight. This panel explores where agents are already delivering, the architecture and guardrails that make them safe to deploy, and the uncomfortable question of accountability when software acts on its own.

Key discussion threads

• From copilots to autonomous and multi-agent systems: what genuinely works in production today.
• Highest-value early use cases: customer service, finance ops, software engineering, back-office automation.
• Orchestration, observability, testing and "human-in-the-loop" vs "human-on-the-loop".
• The governance gap: identity, permissions and audit trails for non-human actors.
• Liability and trust: who owns the outcome when an agent errs — the vendor, the deployer, or the CIO?

Panelists: 

• Igor Babitsch, ServiceNow
• Dragoș Stănescu, CIO, Syngenta, CIO Council Member
• TBC
• TBC
• TBC

Moderator: Bogdan Tudor, Board member, CIO Council Romania

Networking lunch

PANEL 3: AI in Finance & Regulated Industries: Innovating Under the Rules

Banks want to move fast on AI. Regulators, the EU AI Act and DORA say "prove it's safe first." Who wins?

Synopsis

Nowhere is the tension between AI ambition and accountability sharper than in banking and other regulated sectors — a fitting debate to host at the National Bank of Romania. Financial institutions are deploying AI for fraud detection, credit scoring, AML, personalised service and productivity, while operating under intensifying rules: the EU AI Act classes many credit and risk use cases as "high-risk," and DORA raises the bar on operational resilience. With the AI Act's high-risk timeline itself being reshaped by the EU's Digital Omnibus, institutions must innovate against a moving regulatory target. This panel asks how regulated players turn compliance from a brake into a competitive advantage — and where the real red lines should be.

Key discussion threads

• Proven AI value in finance: fraud, AML, credit decisioning, forecasting, customer experience, productivity.
• High-risk AI under the EU AI Act, plus DORA and model-risk expectations — and the shifting Omnibus timeline.
• Explainability, bias and the "computer says no" problem in credit and insurance decisions.
• GenAI on sensitive financial data: data residency, confidentiality and third-party model risk.
• Compliance as differentiator: can trust and governance become a selling point, not just a cost?

Panelists: 

• Razvan Ilievici, Executive Director, Cross IT, Banca Comerciala Romana, CIO Council member
• Yossi Altevet, DeepKeep, CTO and Co-founder
• Laurentiu Mandu, Board Member, CIO Council Romania
• TBC
• TBC

Moderator: Calin Rangu, VicePresident, CIO Council Romania

Title Session:

Synopsys:

Speaker: TBC

Title Session:

Synopsys:

Speaker: TBC

PANEL 4: Sovereign AI & The AI Foundation: Europe's Race for Compute, Energy & Independence

Europe writes the rules for AI but runs on someone else's chips, clouds and models. Can it change that — and at what cost?

Synopsis

AI sovereignty is no longer only about where data is stored. It is about who controls the entire AI foundation — cloud, compute, chips, data, models, infrastructure and energy — and how secure and trusted that foundation is.

Europe is pushing for greater technological autonomy while enterprises remain highly dependent on non-European hyperscalers, chipmakers and foundation models. At the same time, the AI boom is driving unprecedented demand for data centres, GPU capacity and energy, making infrastructure itself a strategic asset.

For CIOs, the question is increasingly practical: where should critical AI workloads run, who controls the underlying technology and data, and how do we keep them secure and trusted?

This panel will explore whether Europe — and Romania — can build the sovereign, scalable and secure AI foundation needed to compete in the global AI economy.

Key discussion threads

• Sovereign AI beyond data residency: control over cloud, compute, data, models and infrastructure.
• The AI Foundation: cloud, GPU/compute capacity, data centres, connectivity and AI-ready data.
• Energy as AI infrastructure: can power availability and grids support the coming compute boom?
• Secure & Trusted AI: cybersecurity, data protection, identity and resilience across the AI stack.
• European strategic autonomy: reducing critical dependencies without sacrificing innovation and scale.
• Romania's opportunity: sovereign cloud, data centres, energy and the potential to become a regional AI infrastructure hub.

Panelists: 

• Konstantinos Stavropoulos, Head of Product Management, AI & Data Analytics Solutions, AUSTRIACARD Holdings 
• Oana Eremia, CIO, MagnaPharm Group, CIO Council member
• Tudor Galos, Senior Consultant, Tudor Galoș Consulting 
• Dan Bogdan, ISG Technology Consulting Manager, Dell Technologies
• TBC

Moderator: Yugo Neumorni, President, CIO Council Romania

CIO Council Awards ceremony II edition