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: 15 records · use cases that answered at least one practice questionReported in placeReported in progressReported not in placeNot applicableOther or ambiguous answerNo answer reported
Pre-deployment testing
Pre-deployment testing
Reported in place: 6 (40.0%)Pre-deployment testing
Reported in progress: 7 (46.7%)Pre-deployment testing
Reported not in place: 1 (6.7%)Pre-deployment testing
No answer reported: 1 (6.7%)
15Reported in place: 6 (40.0%)Pre-deployment testing
Reported in progress: 7 (46.7%)Pre-deployment testing
Reported not in place: 1 (6.7%)Pre-deployment testing
No answer reported: 1 (6.7%)
AI impact assessment
AI impact assessment
Reported in place: 3 (20.0%)AI impact assessment
Reported in progress: 8 (53.3%)AI impact assessment
No answer reported: 4 (26.7%)
15Reported in place: 3 (20.0%)AI impact assessment
Reported in progress: 8 (53.3%)AI impact assessment
No answer reported: 4 (26.7%)
Independent review
Independent review
Reported in place: 3 (20.0%)Independent review
Reported in progress: 5 (33.3%)Independent review
Reported not in place: 1 (6.7%)Independent review
No answer reported: 6 (40.0%)
15Reported in place: 3 (20.0%)Independent review
Reported in progress: 5 (33.3%)Independent review
Reported not in place: 1 (6.7%)Independent review
No answer reported: 6 (40.0%)
Ongoing monitoring
Ongoing monitoring
Reported in place: 4 (26.7%)Ongoing monitoring
Reported in progress: 4 (26.7%)Ongoing monitoring
No answer reported: 7 (46.7%)
15Reported in place: 4 (26.7%)Ongoing monitoring
Reported in progress: 4 (26.7%)Ongoing monitoring
No answer reported: 7 (46.7%)
Operator training
Operator training
Reported in place: 5 (33.3%)Operator training
Reported in progress: 3 (20.0%)Operator training
Other or ambiguous answer: 1 (6.7%)Operator training
No answer reported: 6 (40.0%)
15Reported in place: 5 (33.3%)Operator training
Reported in progress: 3 (20.0%)Operator training
Other or ambiguous answer: 1 (6.7%)Operator training
No answer reported: 6 (40.0%)
Fail-safe
Fail-safe
Reported in place: 6 (40.0%)Fail-safe
Reported in progress: 1 (6.7%)Fail-safe
Not applicable: 7 (46.7%)Fail-safe
No answer reported: 1 (6.7%)
15Reported in place: 6 (40.0%)Fail-safe
Reported in progress: 1 (6.7%)Fail-safe
Not applicable: 7 (46.7%)Fail-safe
No answer reported: 1 (6.7%)
| 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 | 15 | 6 | 7 | 1 | 0 | 0 | 0 | 1 | 93% | 40% of 15 |
| AI impact assessment | 15 | 3 | 8 | 0 | 0 | 0 | 0 | 4 | 73% | 20% of 15 |
| Independent review | 15 | 3 | 5 | 1 | 0 | 0 | 0 | 6 | 60% | 20% of 15 |
| Ongoing monitoring | 15 | 4 | 4 | 0 | 0 | 0 | 0 | 7 | 53% | 27% of 15 |
| Operator training | 15 | 5 | 3 | 0 | 0 | 0 | 1 | 6 | 60% | 33% of 15 |
| Fail-safe | 15 | 6 | 1 | 0 | 7 | 0 | 0 | 1 | 93% | 75% of 8 |
| Appeal process | 15 | 0 | 1 | 0 | 13 | 0 | 0 | 1 | 93% | 0% of 2 |
| User and public consultation | 15 | 6 | 4 | 0 | 0 | 0 | 5 | 0 | 100% | 40% of 15 |
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 | |
|---|---|---|---|---|---|---|---|---|---|
| DOE | 33% | 0% | – | – | 17% | 0% | 0% | 17% | 6 |
| SBA | 33% | 33% | 33% | 33% | 33% | 100% | 0% | 33% | 3 |
| FERC | 100% | 100% | 100% | 100% | 100% | 100% | 0% | 100% | 2 |
| DHS | 0% | 0% | 0% | 0% | 0% | 0% | 0% | 0% | 1 |
| EAC | 0% | – | 0% | – | – | – | – | 100% | 1 |
| DOI | – | – | 0% | 0% | 0% | 0% | 0% | 0% | 1 |
| STATE | 100% | 0% | 0% | 100% | 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
12 records with at least one practice unanswered or in progressOpen in explorer| Use case | Stage | Test | AIA | Review | Monitor | Train | Fail-safe | Appeal | Consult |
|---|---|---|---|---|---|---|---|---|---|
| License Plate Capture and Analysis DHS · DHS-2666 | Deployed | Reported in progress | Reported in progress | Reported in progress | Reported in progress | Other or ambiguous answer | Reported in progress | Reported in progress | Reported in progress |
| LivChat DOE · DOE-613 | Deployed | Reported in place | No answer reported | No answer reported | No answer reported | No answer reported | Not applicable | Not applicable | Other or ambiguous answer |
| Electronic Health Record AI Enhancements STATE · R3248 | Deployed | Reported in place | Reported in progress | Reported in progress | Reported in place | Reported in place | Reported in place | Not applicable | Reported in place |
| OpenAI ChatGPT Pilot DOE · DOE-526 | Pilot | Reported in progress | Reported in progress | No answer reported | No answer reported | No answer reported | Not applicable | Not applicable | Other or ambiguous answer |
| xAI Grok Pilot DOE · DOE-601 | Pilot | Reported in progress | Reported in progress | No answer reported | No answer reported | No answer reported | Not applicable | Not applicable | Other or ambiguous answer |
| ChatSRS DOE · DOE-638 | Pre-deployment | Reported in place | No answer reported | No answer reported | No answer reported | Reported in place | Not applicable | Not applicable | Reported in place |
| Microsoft Copilot Pilot DOE · DOE-676 | Pilot | Reported in progress | Reported in progress | No answer reported | No answer reported | No answer reported | Not applicable | Not applicable | Other or ambiguous answer |
| Anthropic Claude Pilot DOE · DOE-703 | Pilot | Reported in progress | Reported in progress | No answer reported | No answer reported | No answer reported | Not applicable | Not applicable | Other or ambiguous answer |
| CHS Q Business AI Assistant (theKraken) DOI · DOI-0169 | Pilot | No answer reported | No answer reported | Reported in progress | Reported in progress | Reported in progress | Not applicable | Not applicable | Reported in progress |
| Internal Generative AI Tools for Staff Productivity EAC · EAC-2 | Pilot | Reported not in place | No answer reported | Reported not in place | No answer reported | No answer reported | No answer reported | No answer reported | Reported in place |
| AI Incident Summarization SBA · SBA-35 | Pre-deployment | Reported in progress | Reported in progress | Reported in progress | Reported in progress | Reported in progress | Reported in place | Not applicable | Reported in progress |
| AI-Powered Search and Recovery SBA · SBA-41 | Pre-deployment | Reported in progress | Reported in progress | Reported in progress | Reported in progress | Reported in progress | Reported in place | Not applicable | Reported in progress |
Unanswered analytical questions
Generated from the data; for follow-up with agencies- VA designated 10 use cases high-impact but answered none of the minimum-practice questions for any of them. Were the practices assessed, and where is that recorded?
- DOJ designated 6 use cases high-impact but answered none of the minimum-practice questions for any of them. Were the practices assessed, and where is that recorded?
- 2 high-impact use cases are reported as deployed while their AI impact assessment is reported in progress. Was deployment authorised before the assessment was complete?
- 1 high-impact use cases mark an appeal process as not applicable. On what basis was each individual outcome judged not contestable?
- 1 high-impact use cases report using PII but give no link to a Privacy Impact Assessment. Is a PIA published elsewhere?
- 5 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.