Get AI-ready: Developing a new data mindset

Videos and Webinars

The rapid advancement of AI is revolutionizing how organizations manage data and approach governance. As AI becomes increasingly integral to business operations, the distinction between data and records is blurring, necessitating a fresh perspective on information management.

Sue Trombley

August 25, 2025

During our August Education Series webinar, I had the opportunity to meet with two Iron Mountain subject matter experts to explore shifting trends in AI and data governance. I was joined by Hana Laws IGP, Principal Consultant, Information Governance & Digital Solutions, and Brian O’Flynn, Director, Marketing Technology & AI, for this interactive conversation.

Our discussion highlighted how a focus on optimization and the emergence of agentic AI are redefining how we manage data, information, and risk. The rapid evolution of artificial intelligence has moved us beyond the realm of simple experimentation and into a new era of optimization. The conversation has shifted from asking, “what can AI do?” to “what tangible value can AI deliver?” This critical shift is driving organizations to move from pilot projects to full-scale production, demanding a new approach to information and data governance. This is reflected in the serious investment we’re seeing, with IT leaders projecting that 20% of technology budgets in 2025 will be dedicated to AI.

From generative to agentic: The next frontier of AI

Our conversation revolved around three core themes that have emerged in the past year: a strategic focus on optimization and enterprise value; the rapid pace of technological advancement, including reasoning and multimodality; and the complex economic and social impact of AI. The market is now squarely focused on enterprise needs, security, and profitability, with a recent Google Cloud report suggesting that more than 70% of organizations are already seeing a return on investment from their generative AI initiatives.

Perhaps the most significant development, however, is the beginning of the transition from generative AI to agentic AI. While generative AI assists and recommends, agentic AI acts. Using a simple analogy: if an individual AI agent is a single musician, agentic AI is the entire orchestra working together to perform a symphony, with the human as the conductor. This system of multiple agents can collaborate, reason, and execute complex, multi-step workflows with minimal human supervision. Real-world examples are already emerging in fields like finance, where agentic AI is revolutionizing fraud detection by operating autonomously in real time to initiate multi-step workflows to protect the customer and the bank. Similarly, in healthcare, agentic systems are being used to continuously monitor patient vital signs and interpret data to anticipate problems before they become critical.

The conversation has moved beyond basic pattern matching to what we’re calling AI reasoning. This is the capacity for advanced learning and decision-making that is beginning to mimic human cognition.

Brian O’FlynnDirector, Marketing Technology & AI, Iron Mountain

This leads to the critical question of “who’s behind the agents?” and “where is the human in the loop?” While agentic AI will become increasingly autonomous, we need to keep the human at the top, making key decisions. This involves building in supervisor agents and compliance agents that can self-criticize and self-correct their own workflows.

The convergence of governance and the challenge of retention

The rise of agentic AI necessitates a new mindset for information governance, moving away from siloed practices. There’s an increasing convergence of information, data, and AI governance. Information governance focuses on compliance and records retention, data governance on data quality and usability, and AI governance on ethical risks and business alignment. The new paradigm calls for these areas to work collaboratively, with governance becoming an “AI native” engine that drives trust and automation at scale.

We’re going to see AI really featuring massively and more predominantly in the ability to automate the data and information governance.

Hana LawsIGP, Principal Consultant, Information Governance & Digital Solutions, Iron Mountain

We also addressed a fundamental question “Do we keep data forever?” and agreed on the importance of moving beyond the traditional records-retention mindset. Data retention policies must now be explicitly tied to the AI model lifecycle, requiring organizations to retain training data for debugging, audits, and compliance. At the same time, businesses must carefully assess the strategic value of their historical data, recognizing that not all data is equally valuable. We need to define the purpose of data, understand applicable regulations, and establish new records and data and information retention policies that are clear and well-defined.

How to stay ahead of the curve

With legislation still catching up to the pace of AI advancement, organizations are left to navigate a complex regulatory landscape that varies across the globe. The EU is taking a comprehensive approach with the EU AI Act, while the US is following a more piecemeal, state-led path.

Here’s some practical advice for professionals to stay informed and manage this rapid change:

Curate your sources: Subscribe to a small number of high-quality newsletters and podcasts to avoid being overwhelmed by the sheer volume of information.
Follow thought leaders: Use platforms like LinkedIn to follow a handful of respected AI experts to cut through the noise.
Test and experiment: Actively explore the new functionalities of AI-enabled tools in your personal and professional life.
Leverage specialized resources: Use tools like the IAPP Global AI Law and Policy tracker for up-to-date information on legislation and compliance.
Encourage internal dialogue: Build forums or host brown-bag lunches where colleagues can share insights and improve the organization’s overall AI literacy.

The shift to agentic AI and its profound impact on governance and data management presents both immense challenges and opportunities. By moving from a mindset of experimentation to one of optimization, and by converging our governance practices to reflect this new reality, organizations can orchestrate their data safely and securely to meet the future head-on.

Watch the full webinar

Interested in learning more about this topic and hearing the live Q&A with our panelists? Visit Iron Mountain’s 2025 Education Series to watch the on-demand recording of Get AI-ready: Developing a new data mindset.