Governance, Skills, Integration: CIO Responsibilities for Agentic AI – an interview with Levent Ergin (Informatica)
Agentic AI is redefining enterprise data management – but without the right foundation, its potential will remain untapped. Levent Ergin is Chief Strategist for Climate, Sustainability and Artificial Intelligence at Informatica. Discover how Informatica’s Intelligent Data Management Cloud can turn chaos into business value faster, with data everyone can trust, supporting enterprise use cases with AI-powered cloud data management.
In this Confare interview, Levent shares practical insights for CIOs looking to make agentic AI work at scale – from modernizing enterprise data infrastructure to aligning governance, security, and skills across the organization.
Informatica is a proud partner of the Confare CIOSUMMIT Zurich 2025. Connect with peers and industry leaders driving the next evolution of data and AI in the enterprise.
What foundational changes must CIOs prioritize in their enterprise data architecture to effectively implement and scale agentic AI solutions?
As CIOs lead the way in implementing and scaling agentic AI solutions, they must prioritize foundational changes such as modernizing data architecture with scalable, metadata driven cloud-native infrastructures and real-time processing capabilities. Strengthening data quality, governance, and security while ensuring regulatory compliance is essential, alongside enabling seamless integration across diverse sources. Supporting the entire AI model lifecycle and investing in observability tools fosters transparency and trust. Ultimately, cultivating a data-driven culture with skilled talent is critical to unlocking the full potential of autonomous AI systems.
How can CIOs ensure robust data governance, security, and ethical compliance when deploying autonomous AI agents handling sensitive enterprise data?
When it comes to making sure data governance, security, and ethical compliance are solid when using autonomous AI agents with sensitive enterprise data, CIOs need to have robust metadata management rules in place about who can access what data and how it’s used. It’s critical to be aware of the various AI Regulations (EU AI Act, SDAIA, etc.) which may impact them and the legal entities which operate in those jurisdictions, as this will guide the AI Governance Frameworks which need to be implemented across the group.
It’s also really important to take a zero-trust approach—so no one gets in without proper checks, data is encrypted, and there’s constant monitoring to catch any unusual activity right away. Ethical compliance means keeping an eye on the AI’s decisions to make sure they’re fair and don’t cause unintended harm, which means using tools that make AI actions more transparent. And most importantly, CIOs have to work closely with legal teams, IT, and business leaders to keep everything aligned with laws and company values. That teamwork helps build trust in the AI from start to finish.
Join Informatica and other industry leaders at the Confare CIOSUMMIT Zurich 2026 to explore the next steps in data-driven, AI-powered enterprise transformation.
What strategies should CIOs adopt to address skill gaps and foster collaboration between data professionals and AI-driven agentic data management systems?
By combining good leadership, learning, teamwork, and the right tools, CIOs can set their teams up for success with agentic AI. CIOs need to lead the charge by investing in ongoing training to help their data teams stay current with the latest AI and data tools.
In addition to this various AI Regulations (EU AI Act, SDAIA, etc.) will require ‘Risk-based classification’ which requires a cross functional governance, risk & compliance (GRC) approach, where data professionals play a key role. They should bring together people from data, AI, and business areas so everyone can work closely and understand each other’s needs. It’s also important to create a culture where trying new things and sharing what’s learned—both wins and mistakes—is encouraged. Plus, CIOs should support the use of easy-to-use AI tools that let data pros get value from AI without needing to be experts overnight.
In what ways can CIOs leverage agentic AI to not only automate routine data tasks but also accelerate strategic decision-making and innovation in their organizations?
CIOs can get a lot more out of agentic AI than just automating boring, routine tasks. Sure, that saves time and cuts down on mistakes, but the real benefit comes when AI digs into complex data and offers smart advice that helps leaders make quicker, better decisions.
On top of that, agentic AI can uncover patterns or opportunities that people might overlook, which can lead to new ideas and innovation. By using AI not just to handle the small stuff but to boost big-picture thinking, CIOs can keep their organizations flexible and ahead of the game. Having said that, just because a task can be automated, it may not necessarily mean it should be, as taking a risk-based approach may reveal that making mistakes could have a huge societal impact, for example in the case of a bank its ATM network provides a critical economic function and may not warrant such automation, whereas other functions within that same bank may benefit hugely.
What are the critical challenges CIOs should anticipate when integrating agentic AI with existing enterprise systems, and how can they prepare their teams and infrastructure to overcome them?
When CIOs bring agentic AI into their current systems, they should expect a few challenges. One big one is making sure the AI plays nice with older systems that weren’t built for this kind of tech, where metadata plays such an important role in modernizing legacy technology for AI. Data can also be messy, biased, or inconsistent, which makes it hard for AI to do its job well, due to Garbage In Garbage Out (GIGO). Taking a risk-based approach is critical. Processes which have clearly defined boundaries with low risk are great to start as an MVP, starting small and scaling from a foundation of success. Plus, teams might need some extra training and support to get comfortable working with AI.
To get ahead of these issues, CIOs should focus on updating their infrastructure to be more flexible, cleaning up their data, and helping their people learn and adjust to the changes. It’s also important to have strong rules around governance and security to keep everything safe. Being ready like this makes it easier to get the most out of agentic AI.