Artificial Control and the rise of agentic AI

With Michael Savoie, PhD – Co-Founder, CyberReady, LLC

Artificial control: a deep dive into the evolving principles of AI Governance using the commercial real estate (CRE) industry as an example.

There has been a persistent trend in the last few years to think of artificial intelligence as a high-end amenity – something that’s “nice to have”. But the power of AI continues to evolve, opening up new opportunities for us. So, we must ask ourselves the pivotal question: who is really in control? And how do we organize AI governance?

In 2025, I, along with Michael Savoie of CyberREADY, LLC, wrote an in-depth article for Summit Journal, where we unpack the modern challenges that come with the next phase of the AI revolution. 

We noted that, although it still provokes a lot of debate, industries are rapidly becoming comfortable with the use of generative AI (ChatGPT, Gemini, etc). Using CRE as an example, these systems can be extremely helpful for things like summarising leases and drafting emails. 

But if we look forward, there is an entirely new avenue of AI which carries far more powerful capabilities, and has the potential to bring us into much more volatile territory if we don’t manage it properly: agentic AI.

To understand the difference, think of an orchestra: The generative AI systems that we’re all used to are the musicians in that orchestra. Agentic AI, however, is the conductor, entrusted with much greater responsibilities of coordination, able to command other AI functionalities with a degree of decision-making that we’re probably not used to.  

What does that look like on the ground? In CRE terms, for example, it means that AI systems might not just analyse a tenant’s credit score, but will also manage the greater aspects of approval and contact. The system might predict a potential tenant’s likelihood of renewal, managing its outreach based on those conclusions. Meanwhile, it can also adjust things like facility-management schedules in real-time, giving us an efficiency that we haven’t known before. 

But there are traps. You’ve probably heard of them before, and they centre on the potential flaws inherent to unmanaged artificial intelligence. There’s a “shadow side” to it all, rooted in issues like bias and “hallucinations”. These contribute to a flawed perspective and a systemic danger for business leaders. You can read more about it in the full article. 

Even now, the ground continues to shift beneath entire industries like CRE, which are rapidly adjusting their day-to-day working structures to accommodate the capabilities of artificial intelligence. The intersection between data privacy and AI creates complex compliance challenges as AI systems fundamentally change how personal data is processed, analyzed, and used for decision making, thus amplifying the need for transparency and disclosure from market players. This should sound warning bells in the eyes of corporate leaders, who must ensure that they can embrace this new technology without allowing for a loss of control. 

The old saying maintains that “a computer can never be held accountable, therefore a computer must never make a management decision.” This is crucial to remember in the modern world of evolving AI, and indicative of the structures we need to put in place to ensure we have a positive roadmap for AI’s future. 

Read more here.