AI agents are already interacting with businesses in new ways. Some arrive through APIs. Others browse websites, compare products and complete transactions through the same interfaces customers use. For CIOs, the challenge is no longer simply making data accessible or keeping systems secure. It's deciding how to welcome these new interactions while protecting the business.
The instinct to lock down is understandable. AI agents can introduce risks around data misuse, unauthorized access and fraudulent transactions. But shutting them out entirely carries commercial risks of its own. As consumers delegate more buying decisions to AI, businesses that aren't accessible to agents could miss out on valuable customers and sales.
The answer is governed openness: a deliberate approach to deciding what agents can see, what they can do and under what conditions.
So where should CIOs start? Here's a three-step playbook.
1. Assess. Know where you stand
Before you can decide how to engage with AI agents, you need to understand how they're already interacting with your business.
Run your product data through an AI agent query and see what comes back. Browse your own website the way a browser-based agent would, systematically and without the assumptions a human might make when information is incomplete or unclear.
Then ask some important questions:
- Can agents find and accurately interpret your product information?
- How much of your website activity is already agent-initiated?
- Can your systems distinguish between human and agent interactions?
The goal: Establish a clear picture of your current readiness, including where your visibility and controls fall short.
2. Build. Create an agent-ready foundation
Build the agent-ready data layer that makes your enterprise visible, credible and trustworthy to the agents that consumers are already using. That means structured data, clear provenance, real-time availability, agent detection and an orchestration layer that connects your systems into a coherent, agent-queryable infrastructure.
There's also a strategic decision to make: what should you build, buy or develop with partners? Make that choice deliberately, with governance in mind from the start.
The next 12 to 18 months could be critical. Enterprises that invest in this foundation now have an opportunity to build an advantage that compounds over the following three years. Those who wait will risk spending significantly more to catch up.
The goal: Build a connected, trustworthy foundation that supports agent interactions and creates a lasting competitive advantage.
3. Govern. Set the rules of engagement
Design an agent governance framework that makes your openness to AI agents sustainable, both commercially and reputationally.
That means authentication, access control, agent detection, audit trails and monitoring, along with clear escalation protocols that bring a human into the loop when an autonomous interaction produces an unintended outcome.
This also requires collaboration between the CIO, CMO and commercial teams to ensure governance reflects both technical safeguards and business priorities.
The goal: Make agent interactions safe, accountable and commercially valuable.
“Agent governance isn't a constraint on this new era of commerce. It's the foundation that makes it trustworthy enough to scale.”
Dirk Jan van der Pol, Head of Commerce, Empathy Lab
The opportunity is already here
The agents are already knocking. Some are at your API door; some are already inside. They are browsing your website, reading your content and evaluating your brand with a systematic thoroughness that few human shoppers can match.
The enterprises that recognize these different interactions, serve them deliberately and govern them carefully will be best placed to turn AI agents into one of their most valuable new routes to customers.
The window to get this right is open. It will not stay open indefinitely.
Ready for this shift? Find out where you stand.
Before deciding which agents to welcome, what to protect and where to invest, you need a clear picture of your readiness.
Start with our free Agentic Commerce Readiness Starter Assessment for a quick snapshot of where you stand and where to focus next.
For a deeper dive, Empathy Lab works with CIOs and their commercial counterparts to assess agent activity, data and content readiness, and governance gaps, establishing a clear baseline and a prioritized path forward. Get in touch for more information.
Contributors in this article