Governance and accountability
Governance evidence
What agencies reported about the OMB M-25-21 minimum practices for high-impact AI. This is a disclosure analysis, not an audit, a compliance finding or an official assessment. An unanswered question is missing evidence, not evidence that a safeguard is absent.
Practice-by-practice evidence
Denominator: 69 records · use cases that answered at least one practice questionReported in placeReported in progressNot applicablePrecluded by law or guidanceWaived by CAIOOther or ambiguous answerNo answer reported
Pre-deployment testing
Pre-deployment testing
Reported in place: 64 (92.8%)Pre-deployment testing
Reported in progress: 2 (2.9%)Pre-deployment testing
No answer reported: 3 (4.3%)
69Reported in place: 64 (92.8%)Pre-deployment testing
Reported in progress: 2 (2.9%)Pre-deployment testing
No answer reported: 3 (4.3%)
Independent review
Independent review
Reported in place: 67 (97.1%)Independent review
Reported in progress: 1 (1.4%)Independent review
Waived by CAIO: 1 (1.4%)
69Reported in place: 67 (97.1%)Independent review
Reported in progress: 1 (1.4%)Independent review
Waived by CAIO: 1 (1.4%)
Ongoing monitoring
Ongoing monitoring
Reported in place: 62 (89.9%)Ongoing monitoring
Reported in progress: 4 (5.8%)Ongoing monitoring
No answer reported: 3 (4.3%)
69Reported in place: 62 (89.9%)Ongoing monitoring
Reported in progress: 4 (5.8%)Ongoing monitoring
No answer reported: 3 (4.3%)
Operator training
Operator training
Reported in place: 64 (92.8%)Operator training
Reported in progress: 1 (1.4%)Operator training
Other or ambiguous answer: 1 (1.4%)Operator training
No answer reported: 3 (4.3%)
69Reported in place: 64 (92.8%)Operator training
Reported in progress: 1 (1.4%)Operator training
Other or ambiguous answer: 1 (1.4%)Operator training
No answer reported: 3 (4.3%)
Fail-safe
Fail-safe
Reported in place: 47 (68.1%)Fail-safe
Not applicable: 19 (27.5%)Fail-safe
No answer reported: 3 (4.3%)
69Reported in place: 47 (68.1%)Fail-safe
Not applicable: 19 (27.5%)Fail-safe
No answer reported: 3 (4.3%)
| Practice | In scope | In place | In progress | Not in place | N/A | Precluded or waived | Other | No answer | Answered | In place of applicable |
|---|---|---|---|---|---|---|---|---|---|---|
| Pre-deployment testing | 69 | 64 | 2 | 0 | 0 | 0 | 0 | 3 | 96% | 93% of 69 |
| AI impact assessment | 69 | 69 | 0 | 0 | 0 | 0 | 0 | 0 | 100% | 100% of 69 |
| Independent review | 69 | 67 | 1 | 0 | 0 | 1 | 0 | 0 | 100% | 99% of 68 |
| Ongoing monitoring | 69 | 62 | 4 | 0 | 0 | 0 | 0 | 3 | 96% | 90% of 69 |
| Operator training | 69 | 64 | 1 | 0 | 0 | 0 | 1 | 3 | 96% | 93% of 69 |
| Fail-safe | 69 | 47 | 0 | 0 | 19 | 0 | 0 | 3 | 96% | 94% of 50 |
| Appeal process | 69 | 19 | 1 | 0 | 46 | 3 | 0 | 0 | 100% | 95% of 20 |
| User and public consultation | 69 | 34 | 0 | 0 | 0 | 0 | 32 | 3 | 96% | 49% of 69 |
Answered = records with any answer ÷ records in scope. In place of applicable = reported in place ÷ (in scope minus not applicable, precluded and waived). Unanswered records stay in that denominator, so the rate is a lower bound on reported evidence, not an estimate of real practice.
| Testing | Impact assess. | Indep. review | Monitoring | Training | Fail-safe | Appeal | Consultation | n | |
|---|---|---|---|---|---|---|---|---|---|
| DHS | 100% | 100% | 100% | 100% | 100% | 100% | 100% | 62% | 26 |
| SBA | 100% | 100% | 100% | 100% | 100% | 100% | 100% | 65% | 17 |
| DOE | 82% | 100% | 100% | 82% | 91% | 100% | 100% | 18% | 11 |
| SSA | 100% | 100% | 100% | 88% | 100% | 100% | 100% | 25% | 8 |
| FERC | 40% | 100% | 100% | 40% | 40% | 40% | 0% | 40% | 5 |
| DOL | 100% | 100% | 100% | 100% | 0% | 0% | 100% | 0% | 1 |
| TREAS | 100% | 100% | 0% | 0% | 100% | 100% | 0% | 100% | 1 |
Hatched cells: the agency answered nothing for that practice, so there is no evidence either way. Compare agencies with care: each agency is one reporting unit, and portfolios differ in size and type.
Records requiring further investigation
8 records with at least one practice unanswered or in progressOpen in explorer| Use case | Stage | Test | AIA | Review | Monitor | Train | Fail-safe | Appeal | Consult |
|---|---|---|---|---|---|---|---|---|---|
| Copilot Studio DOE · DOE-444 | Deployed | Reported in progress | Reported in place | Reported in place | Reported in progress | Reported in place | Reported in place | Not applicable | Other or ambiguous answer |
| Copilot for Microsoft 365 DOE · DOE-452 | Deployed | Reported in progress | Reported in place | Reported in place | Reported in progress | Reported in place | Reported in place | Not applicable | Other or ambiguous answer |
| GitHub Copilot DOE · DOE-586 | Deployed | Reported in place | Reported in place | Reported in place | Reported in place | Reported in progress | Reported in place | Not applicable | Reported in place |
| GenAI Productivity Tool TREAS · TREASURY-OCIO-001 | Deployed | Reported in place | Reported in place | Reported in progress | Reported in progress | Reported in place | Reported in place | Reported in progress | Reported in place |
| Improve Safety Inspections Use Case FERC · FERC-0002 | Pre-deployment | No answer reported | Reported in place | Reported in place | No answer reported | No answer reported | No answer reported | Not applicable | No answer reported |
| Enhance Market Surveillance and Fraud Detection Use Case FERC · FERC-0003 | Pre-deployment | No answer reported | Reported in place | Reported in place | No answer reported | No answer reported | No answer reported | Not applicable | No answer reported |
| Support Interconnection Request Responses Use Case FERC · FERC-0004 | Pre-deployment | No answer reported | Reported in place | Reported in place | No answer reported | No answer reported | No answer reported | Not applicable | No answer reported |
| Policy Assistance Tool (PAT) SSA · R3116 | Pilot | Reported in place | Reported in place | Reported in place | Reported in progress | Reported in place | Reported in place | Not applicable | Reported in place |
Unanswered analytical questions
Generated from the data; for follow-up with agencies- 24 high-impact use cases mark an appeal process as not applicable. On what basis was each individual outcome judged not contestable?
- 10 high-impact use cases report using PII but give no link to a Privacy Impact Assessment. Is a PIA published elsewhere?
- 20 governance answers are ambiguous (for example, an answer option from a different question). What did the agency intend?
These are questions the inventory raises but cannot answer. They are not findings of non-compliance.