14 Human-in-the-Loop Patterns for AI Agents

TL;DR Human-in-the-loop AI agents hand some decisions to a person, such as approving a payment before it is sent. We studied AI regulations, our customers, our own systems, and published applications, and found 206 human controls across 225 regulations. Almost all of them fit one of 14 patterns, each set by who decides, the deadline, and what happens if nobody decides. Sections IRules beyond the wait IISources and method IIIThe oversight event and its settings IVPattern 1 - Action approval VPattern 2 - Decision review VIPattern 3 - Change authorization VIIPattern 4 - Exception VIIIPattern 5 - Match and adjudication IXPattern 6 - Outcome reconciliation XPattern 7 - Contest and appeal XIPattern 8 - Override XIIPattern 9 - Flag and feedback XIIIPattern 10 - Triage XIVPattern 11 - Hand-off XVPattern 12 - Halt and restart XVIPattern 13 - Attestation XVIIPattern 14 - Oversight quality XVIIIReviewer screens XIXImplementation XXLessons learned XXILimits of the evidence XXIIConclusion XXIIIAppendix A: Human controls in the corpus by pattern XXIVAppendix B: The decision vocabulary Every AI system we have helped build reaches a point where a person has to decide. A payments agent matches an invoice to a purchase order and asks to pay the supplier 4,200 EUR. A notices agent drafts an adverse decision letter to a loan applicant. A model team proposes sending a new model 10% of production traffic. The payment, the letter and the rollout wait for a person to decide before the work continues. Our customers build each human decision with the Akka SDK. A workflow step pauses for the person and resumes when the decision arrives. A timeout on the pause can send a reminder or move the decision to someone more senior. An autonomous agent can hand a task to a person and wait for the result. The workflow and the task are durable, so a wait that lasts days survives a restart of the service. Until now, each of these human processes was designed individually, as its own Akka workflow. The team building the application decided who may answer, how long the answer may take, what happens when nobody answers, and what record the answer leaves. Here are some examples in real-world AI systems we have helped build where a human-in-the-loop control existed: A beverage manufacturer's data-use governance system sends every recommendation to a dataset owner or the governance lead, who confirms, overrides or escalates it with a mandatory reason. The reminder after three days and the escalation after seven are application code around a workflow pause. A medical-supply company's voice agent routes a call to a person at 21 points in its four call flows. We did the same thing in our own services. A loan origination service, a trace ledger and a context layer, all built at Akka, each implemented the same oversight contract by hand. The loan origination service and the trace ledger marked a decision at risk when 75% of its time had passed, and the context layer marked it at 80%. When we looked in our own regulatory corpus in Akka Verify, we found 206 controls that require a person to act. A control states one obligation from a regulation or framework and names the mechanism that enforces it. Our own schema had no way to describe what that person decides. We also read the documentation of 34 products in AI governance, LLM operations and cloud agent services. Only three AI governance platforms among them turn a monitoring signal into a work item that someone owns and closes with a recorded outcome. Across our customers' systems, our own services and the regulations, the same human processes kept recurring. We found fourteen recurring patterns, and each one is described below with the obligation behind it, its settings and a working example. Each pattern raises an oversight event, the record of one decision a person must make, and every oversight event moves through one set of states. The patterns differ in who decides, the deadline, the outcome when nobody decides, the allowed decisions and the record kept. The fourteen patterns will soon be available as templates in the Akka Agentic AI Platform. Pattern When the person acts What happens 1. Action approval Before the system acts A person approves an action before the system takes it. 2. Decision review Before the system acts A person reviews a decision about someone before it takes effect. 3. Change authorization Before the system acts Two people who neither proposed nor built a change approve it before it reaches production. 4. Exception Before the system

Excerpt shown under the publisher's RSS terms · 200 words · full text and figures are on the publisher's site