Show the proof. Name the limit.
TELOS claims are bounded to the source, test, runtime path, and artifact opened for review. A passing proof says what happened on that path. It does not turn one integration into a universal product claim.
What current evidence supports
Framework-native observation
TELOS has a runnable Strands proof that observes the framework's real agent and tool loop, correlates action IDs, and emits local receipts without changing the tool result.
Bound to the tested Strands source commit and local harness. This is integration evidence, not proof of coverage across every agent framework.
Hash-linked receipt integrity
The tested receipt path links records and rejects a tampered historical entry during verification.
Hash linkage demonstrates integrity behavior inside the tested chain. It is not the same as live cryptographic signing or third-party attestation.
Passive action scoring
The current internal deployment computes and records governance verdicts while allowing the underlying action to proceed.
Passive scoring supports observation and disclosure. It does not prove blocking, least-privilege restriction, or complete input and output capture.
What current evidence does not support
- Universal detection of every departure or unsafe action
- Universal capture of all model inputs and outputs
- Blocking or enforcement in the current internal passive deployment
- Live cryptographic signing of every operational event
- Production readiness across agent frameworks or customer environments
- Certification, regulatory approval, or blanket conformance
- Customer counts, commercial pricing, or outcome guarantees
Verification method
Positive proof
- Pin the source or artifact under review
- Run the real integration path
- Preserve exact command output and hashes
- State the tested scope
Negative control
- Tamper with a prior receipt
- Remove required provenance
- Test a path outside the claimed integration
- Keep the resulting limitation visible
Mapped to eleven governance frameworks, self-assessed and not independent certifications.
TELOS maintains alignment mappings from its observation and attestation work to eleven regulatory and standards frameworks, with article and control anchors. These alignment mappings are self-assessed, not independent certifications.
| # | Framework | Year | Scope | Anchors |
|---|---|---|---|---|
| 1 | IEEE 7000 | 2021 | Ethical system architecture | IEEE 7000-2021 |
| 2 | IEEE 7001 | 2021 | Algorithmic transparency | IEEE 7001-2021 |
| 3 | IEEE P7002 | 2022 | Data privacy engineering | IEEE P7002-2022 |
| 4 | IEEE P7003 | 2024 | Algorithmic bias | IEEE P7003-2024 |
| 5 | SAAI Framework | 2026 | Agentic AI safety (Watson & Hessami) | SAAI categories |
| 6 | EU AI Act (2024/1689) | 2024 | AI risk classification and obligations | Art. 9, 11, 12, 14, 15, 72, 73 |
| 7 | NIST AI 600-1 | 2024 | Generative-AI risk management | 12 GAI risks; GV / MAP / MS / MG |
| 8 | NIST AI RMF 1.0 | 2023 | AI risk lifecycle | Govern / Map / Measure / Manage |
| 9 | UC Berkeley CLTC | 2026 | Agentic AI risk profile (NIST RMF) | Govern, Map, Manage subcategories |
| 10 | NAIC Model Bulletin | 2023 | Insurance AI governance | Bulletin §3; Model #880 |
| 11 | OWASP Agentic Top 10 | 2026 | Agent security risks | ASI01-ASI10 |
These are self-assessed alignment mappings maintained in the TELOS governance review repository, not independent certifications. The whitepaper maps each instrument article by article in Regulatory Alignment Map. Mappings are being published progressively. TELOS governs alignment to declared purpose, not correctness of outputs, and does not certify legal compliance.
A complementary layer.
Identity and permission frameworks such as ATF establish who an agent is and what it may do. The Record of Trust sits above that foundation and attests what the agent actually did. The comparison below describes only where our own work sits; it does not speak for those projects.
| Framework | Its lane | Relation to the Record of Trust |
|---|---|---|
| ATF | Zero-Trust front door: who the agent is, what it may do | Attests identity and permission. The record attests the action that followed. |
| AARM | Runtime enforcement layer | Acts on actions at runtime. The record observes and attests; it does not take control. |
| C2PA | Provenance for media content | Attests where content came from. The record attests what an agent did. |
| NIST AI RMF | Risk-management lifecycle | Asks for monitoring and accountability. The record supplies that evidence. |
| EU AI Act | Risk classification and obligations | Asks for oversight and documentation. The record is evidence toward both. |