Your AI Agents Need A Chief Operating Officer, Too

Get Covered now runs 16 AI agents across more than 3 million rental units. In Crunchbase News's report, chief operating officer Rick Folgmann writes that the technical questions of building those agents were the easier ones. The harder question, for each agent, was who it reports to. He argues for running agents like staff in a regulated job, with written rules, a defined scope, a manager, and a regular review.

This piece is for decision-makers scaling agentic AI who will have to answer for what an agent sends. Folgmann has more than 25 years of executive leadership experience in the financial and insurance industries. Get Covered is an insurance technology company that provides cost-effective property coverage with just a few clicks. He describes what has worked there: an existing rulebook, autonomy set by workflow, and a named human owner for every agent.

A Chief Operating Officer, Too: what actually changed

A landlord in California has 21 days after a resident moves out to return the security deposit or send an itemized list of deductions. Miss it in bad faith, and the penalty can reach twice the deposit. Folgmann writes that the deadline does not care whether the letter was supposed to come from a person or a piece of software. Either way, someone will answer for it. A bad-faith miss can still cost twice the deposit when software was supposed to send the notice.

Most companies are about to face that same question, and few are ready for it, Folgmann writes. Deloitte’s 2026 State of AI survey found that 23% of companies use agentic AI at least moderately today, and 74% expect to within two years. He says those figures make the case for running agents like regulated staff before that wider use arrives. He would give the same advice to any decision-maker scaling this work. The preparation is written rules, a defined scope, a manager, and a regular review.

A Chief Operating Officer, Too: how it works

Folgmann starts with tasks whose rules already exist outside anyone’s head: a statute, a contract clause, or an insurance requirement in a vendor agreement. Vendor certificates of insurance are the example. Across the industry, property teams spend roughly three hours per property every week chasing those documents by email. The work is tedious, and it is completely specified. Decision-makers get two advantages. The agent has something objective to check its output against, and the team can audit the result without debating judgment calls.

Folgmann says autonomy is not a fixed property of the product, though executives he speaks with often ask whether an AI system is autonomous. The same agent should run with different levels of freedom depending on what it touches. Get Covered configures each agent on different levels. A reminder to a vendor whose certificate expires next month can go out automatically. A response to a resident who has requested an emotional support animal never does. He tells leaders to list every workflow an agent will touch and ask, “If this action is wrong, how long until we find out, and what will it have cost by then?” Quick detection and low cost earn an agent more room.

A Chief Operating Officer, Too: why it matters now

Gartner has predicted that more than 40% of agentic AI projects will be canceled by the end of 2027. Folgmann says most agent projects that fail will do so because nobody defined the value or controlled the risk, and both of those are management problems. The models, he writes, will keep improving. A company’s clarity about who decides what will not improve on its own. For decision-makers heading into 2027, that clarity is the constraint he says will not fix itself.

McKinsey’s most recent State of AI research found that the small group of companies seeing significant bottom-line returns from AI, about 6% of respondents, were nearly 3x as likely to have fundamentally redesigned their workflows. In Folgmann’s experience, that redesign starts with writing the rule down. If a new hire could not follow a description of a decision on the person’s first day, an agent will not follow it either. He cites that 6% group because the returns showed up alongside a redesigned workflow.

A Chief Operating Officer, Too: who is affected

Earlier in his career, Folgmann learned that every automated credit decision eventually had to be explained to a customer, a regulator, or both. “The model did it” was never an acceptable answer. He says that still holds, except agents now act more often and across more systems than any underwriting model did. The customer and the regulator can still demand the explanation. A named person has to give it, because the model was never an acceptable answer on its own.

Every agent at Get Covered has a person ultimately responsible, the same way a team or a budget does. That person reviews its error log, decides when to widen or narrow its autonomy, and answers for it. Ownership also heads off a quieter problem. When several teams launch agents independently, nobody sees where those agents contradict one another. Folgmann says the named owner is the one who makes that contradiction visible across teams that would otherwise launch agents on their own.

A Chief Operating Officer, Too: what to watch

For decision-makers heading into 2027, Folgmann’s advice is to treat each agent as a new employee in a regulated role. He tells them to write down its rules, limit its scope, and assign a manager who checks its work and discusses its performance. He wants that manager named and in place as companies head into 2027, the horizon of Gartner’s cancellation forecast. Defining the value and controlling the risk, he says, are the management tasks that decide whether those projects survive.

Companies that do this will be able to give agents more responsibility over time, because they will have the evidence to justify it. Businesses that skip it will spend 2027 explaining why their software did something nobody approved. Folgmann says the evidence is the manager’s record: the written rules, the limited scope, the error log, and each decision to widen or narrow autonomy. Get Covered is already running that pattern with 16 AI agents across more than 3 million rental units.