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: 14 records · deployed high-impact use casesReported in placeReported in progressNot applicableOther or ambiguous answerNo answer reported
Pre-deployment testing
Pre-deployment testing
Reported in place: 7 (50.0%)Pre-deployment testing
Reported in progress: 1 (7.1%)Pre-deployment testing
No answer reported: 6 (42.9%)
14Reported in place: 7 (50.0%)Pre-deployment testing
Reported in progress: 1 (7.1%)Pre-deployment testing
No answer reported: 6 (42.9%)
AI impact assessment
AI impact assessment
Reported in place: 4 (28.6%)AI impact assessment
Reported in progress: 4 (28.6%)AI impact assessment
No answer reported: 6 (42.9%)
14Reported in place: 4 (28.6%)AI impact assessment
Reported in progress: 4 (28.6%)AI impact assessment
No answer reported: 6 (42.9%)
Independent review
Independent review
Reported in place: 4 (28.6%)Independent review
Reported in progress: 4 (28.6%)Independent review
No answer reported: 6 (42.9%)
14Reported in place: 4 (28.6%)Independent review
Reported in progress: 4 (28.6%)Independent review
No answer reported: 6 (42.9%)
Ongoing monitoring
Ongoing monitoring
Reported in place: 7 (50.0%)Ongoing monitoring
Reported in progress: 1 (7.1%)Ongoing monitoring
No answer reported: 6 (42.9%)
14Reported in place: 7 (50.0%)Ongoing monitoring
Reported in progress: 1 (7.1%)Ongoing monitoring
No answer reported: 6 (42.9%)
Operator training
Operator training
Reported in place: 6 (42.9%)Operator training
Reported in progress: 2 (14.3%)Operator training
No answer reported: 6 (42.9%)
14Reported in place: 6 (42.9%)Operator training
Reported in progress: 2 (14.3%)Operator training
No answer reported: 6 (42.9%)
Fail-safe
Fail-safe
Reported in place: 6 (42.9%)Fail-safe
Reported in progress: 1 (7.1%)Fail-safe
Not applicable: 1 (7.1%)Fail-safe
No answer reported: 6 (42.9%)
14Reported in place: 6 (42.9%)Fail-safe
Reported in progress: 1 (7.1%)Fail-safe
Not applicable: 1 (7.1%)Fail-safe
No answer reported: 6 (42.9%)
| 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 | 14 | 7 | 1 | 0 | 0 | 0 | 0 | 6 | 57% | 50% of 14 |
| AI impact assessment | 14 | 4 | 4 | 0 | 0 | 0 | 0 | 6 | 57% | 29% of 14 |
| Independent review | 14 | 4 | 4 | 0 | 0 | 0 | 0 | 6 | 57% | 29% of 14 |
| Ongoing monitoring | 14 | 7 | 1 | 0 | 0 | 0 | 0 | 6 | 57% | 50% of 14 |
| Operator training | 14 | 6 | 2 | 0 | 0 | 0 | 0 | 6 | 57% | 43% of 14 |
| Fail-safe | 14 | 6 | 1 | 0 | 1 | 0 | 0 | 6 | 57% | 46% of 13 |
| Appeal process | 14 | 3 | 1 | 0 | 4 | 0 | 0 | 6 | 57% | 30% of 10 |
| User and public consultation | 14 | 4 | 3 | 0 | 0 | 0 | 1 | 6 | 57% | 29% of 14 |
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% | 80% | 80% | 100% | 80% | 100% | 100% | 60% | 5 |
| VA | – | – | – | – | – | – | – | – | 5 |
| FDIC | – | – | – | – | – | – | – | – | 1 |
| DOJ | 0% | 0% | 0% | 0% | 0% | 0% | 0% | 0% | 1 |
| EPA | 100% | 0% | 0% | 100% | 100% | 0% | 0% | 0% | 1 |
| STATE | 100% | 0% | 0% | 100% | 100% | 100% | 100% | 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
10 records with at least one practice unanswered or in progressOpen in explorer| Use case | Stage | Test | AIA | Review | Monitor | Train | Fail-safe | Appeal | Consult |
|---|---|---|---|---|---|---|---|---|---|
| Verification Match Model DHS · DHS-2384 | Deployed | Reported in place | Reported in progress | Reported in progress | Reported in place | Reported in progress | Reported in place | Reported in place | Reported in progress |
| Fingerprint (Friction Ridge) Optical Character Recognition (OCR) DOJ · DOJ-0296 | Deployed | Reported in progress | Reported in progress | Reported in progress | Reported in progress | Reported in progress | Reported in progress | Reported in progress | Reported in progress |
| RCRA PA EPA · 78 | Deployed | Reported in place | Reported in progress | Reported in progress | Reported in place | Reported in place | Not applicable | Not applicable | Reported in progress |
| Automatically-Scored Writing Assessment (AWA) FDIC · FDIC – 32 | Deployed | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
| AI-Augmented Declassification Review STATE · R3213 | Deployed | Reported in place | Reported in progress | Reported in progress | Reported in place | Reported in place | Reported in place | Reported in place | Reported in place |
| VA.gov Chatbot: Summative Content VA · VA-24-2717 | Deployed | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
| Payment Redirect Fraud (PRF) Model VA · VA-24-3291 | Deployed | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
| Automated Decision Support VA · VA-24-4722 | Deployed | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
| Stratification Tool for Opioid Risk Mitigation (STORM) VA · VA-24-4152 | Deployed | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
| REACH VET Suicide Risk Prediction and Recovery Engagement VA · VA-24-4234 | Deployed | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
Unanswered analytical questions
Generated from the data; for follow-up with agencies- VA designated 5 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?
- 4 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?
- 4 high-impact use cases mark an appeal process as not applicable. On what basis was each individual outcome judged not contestable?
- 3 high-impact use cases report using PII but give no link to a Privacy Impact Assessment. Is a PIA published elsewhere?
- 2 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.