AI + Paid Media
The Second Administrator Approved the Change. The Agent Had Already Changed the Target.

Updated: November 15, 2026
Google Ads API v26.2 adds STALE_MULTI_PARTY_AUTH_REQUEST when the target resource was modified or deleted after review began. The approval may be authentic, yet it no longer describes the object the agent is about to change. Store the target resource name, version marker, requested mutation and reviewer decision together. Treat any target change as a reason to stop and rebuild the approval package.
Approval evidence can expire without a timer
Google Ads API v26.2 adds STALE_MULTI_PARTY_AUTH_REQUEST when the target resource was modified or deleted after review began. The approval may be authentic, yet it no longer describes the object the agent is about to change.
Store the target resource name, version marker, requested mutation and reviewer decision together. Treat any target change as a reason to stop and rebuild the approval package.
Compare intent, not only resource identity
A campaign can retain the same ID while budgets, goals, targeting or access have changed. Revalidating only that the resource exists is not enough to preserve the reviewed risk.
Hash or normalize the material fields included in the approval. Before resolution, compare the current object and planned operations with the approved snapshot.
Operating control table
| Control | Evidence | Failure prevented |
|---|---|---|
| Owner | Named reviewer and decision | Unclaimed change |
| Configuration | Saved source state | Cannot reproduce |
| Validation | End-to-end fixture | Partial success accepted |
| Threshold | Success and rollback rule | Result reinterpreted |
| Evidence | Public or reporting read-back | Configured mistaken for delivered |
Separate planning from execution authority
The agent that proposes a sensitive change should not silently broaden it after a human review. Tool retries, optimization passes and conflict resolution can all mutate the plan.
Freeze the approved mutation set. If the agent needs a different resource or operation, create a new request instead of carrying forward the old approval.
Design a clear stale-request response
A stale error should not trigger an automatic retry that recreates the request without explaining what changed. That pattern turns a meaningful control into repeated approval noise.
Show reviewers the before-and-after diff, the reason for the new request and the old decision. Keep the original request closed as evidence rather than overwriting it.
Read back the authorized outcome
After execution, confirm that the final target and permissions match the approved state. A successful mutation proves only that the API accepted it, not that the agent executed the reviewed decision.
Record the resolver, timestamp, target version, response and public or account-level read-back. Escalate any difference before the agent continues to the next dependent action.
Connect the control to the wider system
Use this process alongside the second-administrator control for AI agent access. The platform setting, creative promise, measurement rule and downstream experience should describe the same decision.
Pre-launch checklist
- Write the decision and owner in plain language.
- Capture the source configuration and evidence.
- Test one realistic customer or reporting path end to end.
- Verify mobile presentation, links and accessibility.
- Record the success threshold and rollback trigger.
- Schedule the next review before material spend accumulates.
Run a cross-functional release review
Ask the channel owner, creative lead, analyst and conversion owner to inspect the same case. Each reviewer should state what was configured, what the customer sees, which event proves the outcome and which limitation changes the interpretation.
Repeat the test on mobile and with one adverse condition. Resolve differences in the campaign brief and measurement plan before launch.
Set the review cadence before launch
Assign a monitoring owner, review frequency and evidence threshold before the change is released. Preserve the pre-change baseline and avoid silently editing the definition after results arrive. This turns the launch into a measurable operating decision instead of a one-time interpretation of platform guidance.
At the first review, reconcile the source platform, public experience and one downstream business system. Close the change only when all three support the same conclusion, and record any limitation that future analysts must preserve.
Frequently asked questions
Should this be automated?
Automate repeatable validation and monitoring, but keep material policy, claim and budget decisions inside named approval limits.
What evidence should be preserved?
Keep the source configuration, decision owner, expected outcome, observed result and any exception that changes interpretation.
When should the team roll back?
Rollback when the control is inactive, the customer experience breaks, totals do not reconcile or the agreed threshold is breached.
