Marketing AI agent · consent evidence
Consent & Revocation Evidence AI Agent
Substantiate every send against the record — what the contact agreed to, when the revocation arrived and by what channel — because CASL s. 13 puts the onus of proving consent on the sender.
- Typical delivery
- 4–6 weeks
- Deployment
- Your stack
- Named sender
- Ten-day clock
- After launch
- Agent Care
What this agent does
Assembles the evidence, and nobody signs it
In
- 01
A contact replies STOP, and 64.1200(a)(10) treats that as per se reasonable and starts ten business days.
- 02
A revocation arrives by voicemail, and 64.1200(a)(11) presumes consent gone until the sender rebuts it.
Reason
- 03
A contact clicks unsubscribe, and 15 U.S.C. § 7704(a)(4) closes the same ten business days on the send.
- 04
A Canadian recipient withdraws, and CASL s. 11 wants effect without delay inside ten business days.
- 05
A courier notification is refused, and 64.1200(a)(9)(i)(F) leaves only six business days to honour it.
Decide
- 06
A name goes on the internal list, and 64.1200(d)(6) holds that request for five years.
- 07
A number sits on the national registry, and 64.1200(c)(2) wants a scrub no older than thirty-one days.
Out
- 08
A buyer stops buying, and the relationship exception in 64.1200(f) ends on a seller-specific request.
- 09
Execute write actions only inside the approval boundaries agreed during implementation.
Product statement
The agent captures the consent event and takes the revocation in. A named person decides to send — nothing in this body of law carries the sender signature.
Example workflow
One revocation, receipt to suppression
- 1Revocation evidence receivedReply keywords, call opt-outs, designated forms, voicemail and off-channel email
- 2Revocation context assembledThe contact, the channel it arrived on, the identifiers it reaches and the hour of receipt
- 3Consent evidence draftedThe artefact, its provenance, the revocation and what the file can prove
- 4Controls appliedClock checks, reasonable-method checks, identifier resolution and provenance confidence
No human action required
Stages 1 to 4 run unaided, and nothing is sent at any of them — the agent is assembling, and the compliance lane opens at the completeness gate.
Goes to the named sender to release.
Adds a compliance analyst read first.
Compliance review
The file is held with its revocation, its clock and the identifiers behind it.
Release · Append evidence · Send to compliance
Released — by a named sender
- 6Suppression and contact records updatedOnly where write access and records policy allow it
- 7Outcome evaluatedClock adherence, provenance coverage, analyst corrections and what the read found
Corrections
Each compliance correction is counted in the evaluation.
What should not run autonomously
Human approval stays in control
Outside the boundary — human approval required
8 items
- Deciding that a send may lawfully go out.
- Judging an off-channel revocation rebutted.
- Certifying to a court what consent was held.
- Setting which channels may carry a revocation.
Automation boundary
Agent acts unaided
- ✓Capture the consent event with the provenance that came with it.
- ✓Take revocation in from every channel, then log.
- ✓Run the ten-business-day clock that 64.1200(a)(10) sets running.
- ✓Test each send against what the record can prove.
Nothing is sent or suppressed except by a named person, inside the agreed boundaries.
- Judging whether a consent record would hold up.
- Telling a regulator the estate is compliant.
- Deciding the lawful basis behind a campaign.
- Changes to suppression, consent or contact records.
Example output
One revocation, annotated
Nobody signs a suppression list and no filing exists to make; this record is what one revocation carried into the evidence file.
Evidence file · single revocation
Illustrative example- Revocation
- Reply keyword, 64.1200(a)(10)
- Recorded as
- Honoured within the ten business days
- Channel
- Per se reasonable
- Evidence of record
- Inbound text reply, 3 August 2026
- Confidence
- Held unsent
- Held for
- The releasing sender, by name
As received
Taken from the inbound message log and the consent artefact — it reaches as far as those sources do.
What the record holds
- Inbound message log
- Signed consent artefact
- Identifier match trail
Why no send decision here
No model may decide that a send is defensible; a named person does.
What the score decides
Below the configured threshold a file picks up a compliance read before the sender sees it.
Action
ReleaseAppend evidenceSend to complianceValue
Where AI adds value
The same four claims, placed at the point in the workflow where each one applies.
Where the value lands
Value 01 – 04
Every revocation
From the channel it arrived on
Where the evidence is used
Our audience targeting copilot maintains segments, lookalikes and suppression lists with the permission behind every input on file, so this page assembles no audience and maintains no list — it is revocation intake and the consent evidence file. Its sibling marketing risk assessment agent runs a per-activity assessment under California regulation; here the unit is a per-contact evidentiary record under federal telecom law.
The sender carries the onus
CASL s. 13 says it outright: "A person who alleges that they have consent … has the onus of proving it."
What was checked, and not found
The time-of-day window and call cap widely attributed to Fla. Stat. § 501.059 were absent from two primary fetches of that section; no primary list or count of state mini-TCPA statutes exists; Md. Code Com. Law § 14-3201 only incorporates the federal standard and its remaining provisions were not retrieved; Oklahoma stays unverified; TCPA damages figures were never confirmed in primary text; the suspension of the CASL private right of action rests on secondary sources alone; and 16 CFR Parts 310 and 316 each show last amended 1 August 2026 with that change unchecked.
Build an evidence trail
The contact, the consent captured for them and the hour it was revoked stay together.
Integrations
Typical integrations
Five system groups connect to the same agent. Which of them are in scope is decided in discovery.
Inbound message channels
- SMS reply feeds
- call opt-outs
- Designated forms and numbers
Sending platforms
- Email service providers
- CRM
- Suppression and send logs
Consent capture
- Web forms
- checkout
- lead feeds
- Signed consent artefacts
Agent
Consent and revocation evidence
- Takes the revocation in
- Assembles the evidence
- Holds for the sender
Registry and case systems
- DNC registry
- ticketing
- Scrub and internal-list records
Observability & evaluation
- OpenTelemetry
- Langfuse
- Supported monitoring/evaluation sources
Integration availability depends on the client's existing systems and API access.
Agent controls
Six checks between the model and the sender
Six checks over one contact, and no two alike. Whatever survives them is named in the map below.
L6 · Outermost — last line of defence
Inward → L1 · closest to the model
- L1 – L2
- Test whether a send may stand
- L3
- Puts the release in a person's hands
- L4 – L5
- Keep the contact and the consent behind it
- L6
- Retreats to evidence assembly when signals degrade
How Nestack evaluates it
Evaluate the whole assembly — not only the evidence file that comes out.
Coverage runs the whole depth of the workflow, and every layer is cut by slice.
Surface — the file a court reads
- E1Final-output evaluationDid the file record what the contact actually agreed to?
- E2Step-level evaluationDid the agent read the right contact, the right channel and the live suppression state?
- E3Tool evaluationDid it read and write the correct contact and the correct revocation?
- E4Confidence calibrationDo low-confidence files actually attract more compliance corrections?
- E5Slice evaluationHow does performance change across specific capture channels?
- E6Business outcomeHow many files needed a correction before the sender released?
Floor — the record the sender answers for
Failure modes
Where each failure originates in the agent
Seven failure modes, each fixed at the stage where it first surfaces.
Agent lifecycle
Direction of processing →
01 · Retrieval
1 mode
- KE-03
Stale suppression read
The suppression state read is not the one now live.
Stage gathers
The contact, the consent, the hour and the channel
02 · Reasoning
2 modes
- KE-04
Revocation read too narrowly
An off-channel request is treated as void.
- KE-06
Delayed limb read as live
The cross-channel limb is worked as enforceable.
Stage proposes
The revocations, their clocks and provenance
03 · Tool / write
2 modes
- KE-02
Thin file passed forward
A file moves on without the compliance read.
- KE-05
Bound to the wrong contact
A revocation is filed against the wrong identifier.
Stage writes
Only where write access and approval policy allow it
04 · Output
1 mode
- KE-01
Sent, consent unevidenced
The record shows a send but not what allowed it.
Stage returns
The file a sender releases and a court reads
05 · Change / Version
1 mode
- KE-07
Silent clock regression
A configuration change moves the clock, not the file.
Stage tracks
Model, prompt, clock rules and evidence fields
- Sev-1 · a send after a logged revocation
- Sev-2 · wrong evidence reaches the file
- Sev-3 · source degrades, file held unsent
Affected slices
Imported lists absorb the corrections
A contact-level consent-provenance figure can read clean while imported and partner lists carry most of the rework. Nestack reports the correction rate by capture channel, not only in total.
Slice performance — reported separately, not only in aggregate
Illustrative example| Slice | Failure rate | Lift | Lift vs. threshold | Status |
|---|---|---|---|---|
| Imported and partner lists | 9.7% | 3.6× | Review | |
| Call-centre and voice capture | 6.9% | 2.6× | Review | |
| SMS keyword opt-ins | 4.3% | 1.6× | Watch | |
| Web form and checkout capture | 1.8% | 0.7× | Normal |
Bar: correction-rate lift vs. web form capture baseline · scale 0–4.0× · tick marks the 2.0× review threshold
2 of 4 slices over threshold
Evidence-linked improvement
What an unhonoured revocation costs
A loop ends when the off-channel revocation has become a case the next release must pass. That suite is what the next contact suppressed is measured against.
Improvement cycle · five stages
Switchback — the path turns at Improve and returns at Learn
- 01
Detect
Correction rate rises on imported and partner lists.
- 02
Diagnose
The word STOP typed into a channel nobody was watching, and the send that went anyway, is read back until one cause remains.
- 03
Improve
Number the change; the contacts that drove it are filed underneath it.
- 04
Verify
Each touched contact case is run once more, and one red holds it back.
- 05
Learn
It stays on as a standing test, and the revocation rules move alongside it.
Learn → Detect
The return edge. The next revocation is measured against a suite one case longer.
Typical build scope
Twelve workstreams across six weeks
The build scope read against the delivery timeline. Week structure follows the six-week plan — discovery, sources, evidence assembly, evaluation, integration, then production validation and handover.
Workstream
Week 1Week 2Week 3Week 4Week 5Week 6
- 01Revocation-channel discovery and boundary work.Week 1
- 02Inbound, consent and registry sources.Week 1, Week 2
- 03Revocation-to-clock and provenance-chain mapping.Week 1, Week 2
- 04Revocation and consent ingestion.Week 2
- 05Contact, channel and record binding.Week 2, Week 3
- 06Completeness scoring and review routing.Week 3
- 07Sender release workflow.Week 3, Week 4
- 08Sending-platform integration.Week 5
- 09Revocation and provenance cases.Week 4, Week 5
- 10Guardrails and suppression controls.Week 4
- 11Contact-trail instrumentation.Week 5, Week 6
- 12Deployment, documentation and Agent Care handover.Week 6
12 workstreams · 6 weeks · bar shows the weeks a workstream is active — several run in parallel
Final scope and sequence confirmed in discovery
Engagement tiers
What each tier includes
Rows are the capabilities named in each tier's scope. Higher tiers include everything below them.
| Capability✓ in scope · — not at this tier | PilotOne channel, one send year | ProductionProduction suppression workflow | AdvancedMultiple brands / estates |
|---|---|---|---|
| Introduced at Pilot | |||
| Evidence assembly to your channels | ✓ | ✓ | ✓ |
| Named sender release | ✓ | ✓ | ✓ |
| Consent-provenance baseline | ✓ | ✓ | ✓ |
| Introduced at Production | |||
| Reporting by channel | — | ✓ | ✓ |
| Release workflow in your systems | — | ✓ | ✓ |
| Approved write-back | — | ✓ | ✓ |
| Inbound-message integration | — | ✓ | ✓ |
| Introduced at Advanced | |||
| Multi-brand estates | — | — | ✓ |
| Cross-channel evidence packs | — | — | ✓ |
| Large contact files | — | — | ✓ |
| Multi-clock suppression controls | — | — | ✓ |
| Build price | From $5,000 | From $8,000 | Custom quote |
- Final build price
- Confirmed after discovery based on integrations, workflow complexity, message volume, approval controls and deployment requirements.
- Separate from build
- Build pricing is separate from recurring Agent Care, which covers managed monitoring, evaluations, incidents and verified improvements after launch.
What we need from you
What you bring, and what we build with it
Each input maps to a piece of build scope and a week in the delivery timeline.
You bring
We build with it
01Your sending channels and the estate behind them
→Channel inventory mapping and revocation capture
Week 102Representative consent, revocation and send records
→Record binding, clock logic and the provenance baseline
Week 203Your consent artefacts and where each came from
→Channel mapping, clock binding and the automation boundary
Week 104Access to relevant APIs, feeds or exports
→Inbound, consent and registry source assessment, then integration setup
Week 205Sends you would not want defended
→Revocation cases and the evaluation run
Week 406What no consent record may prove
→Completeness scoring, review routing, guardrails and release controls
Week 307A named sender to release
→Release workflow, then pilot and production validation
Weeks 5–6
Nothing else is required
Deployment, documentation and Agent Care handover are ours.
Delivery timeline
Four phases across six weeks
The widths are the weeks each phase honestly needs, not layout, so a pair has to share the fifth.
Phase
W1
W2
W3
W4
W5
W6
- DiscoveryW1
- BuildW2 – W3
- EvaluateW4 – W5
- Pilot & LaunchW5 – W6
Week focus
W1Consent workflow discovery, channel mapping and the automation boundary
W2Source integration and the consent-provenance baseline
W3Evidence assembly, clock logic and release controls
W4Evaluation suite, revocation cases and failure-mode testing
W5Platform integration, pilot revocations and targeted corrections
W6One send year run under the compliance lead, then Agent Care handover
- Reading the band
- Each bar covers only the weeks its own work is named for. The fifth holds a second because the work does.
- At the end of W6
- When the consent record validates, Agent Care takes the agent on.
- Duration
- Six-week plan shown · typical delivery 4–6 weeks depending on scope confirmed in discovery.
Next step · Marketing AI agent
Build a consent evidence agent around the revocation that landed this morning on a channel nobody watches.
Show us one revocation and the record behind it. The onus of proving consent falls on the sender, never the contact, and since McLaughlin of 20 June 2025 a court reads that record without deferring to any FCC order. Suppression-list upkeep is a different agent.
- Typical delivery
- 4–6 weeks
- Deployment
- Your contacts
- Named sender
- Ten-day clock
- After launch
- Agent Care available
Nestack Agents · consent evidence
AGT-MK-02 · Agent Care available after launch