AI decisions cannot be overturned by humans

Give a human a working way to review and overturn each consequential AI decision: a pending-review step before it takes effect and an override that restores the prior state.

Two pieces on the path where model output becomes an effect (status change, record update, letter, tool execution). Before the effect, consequential decisions wait in a pending_review state or an agent interrupt (LangGraph interrupt_before or interrupt(), Claude Agent SDK can_use_tool) where an assigned overseer sees the output in context and can accept, change, or disregard it. After the effect, an override route (for example POST /decisions/{id}/override or reverse_decision) lets an authorized person reverse the outcome, restores the prior state, and records who overturned it and why, so responsibility is attributable to a person.

Where it goes: 15 agent action surface, 1 application source code, 2 data models.

What reviewers look for: on each path from a model decision to set_status, a record update, a letter, or a tool call, a pending_review queue or interrupt that a person resolves; an override or reversal route that actually restores the earlier state and records the overseer and reason; an assigned overseer role, not an admin-only database edit.

app = builder.compile()
app.invoke({'application': application})

app = builder.compile(checkpointer=MemorySaver(), interrupt_before=['apply_decision'])
config = {'configurable': {'thread_id': application.id}}
app.invoke({'application': application}, config)   # pauses before apply_decision
# overseer UI shows the state; the reviewer accepts, edits, or discards the AI output
app.update_state(config, {'decision': reviewer_decision, 'reviewed_by': reviewer.id})
app.invoke(None, config)

Engineering guidance, not legal advice. Each provision below may add its own details (a cadence, a deadline, a required notice element): open it for those.