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: 12 records · all use cases in the current filterReported in placeReported in progressReported not in placeNot applicableOther or ambiguous answerNo answer reported
Pre-deployment testing
Pre-deployment testing
Reported in place: 2 (16.7%)Pre-deployment testing
No answer reported: 10 (83.3%)
12Reported in place: 2 (16.7%)Pre-deployment testing
No answer reported: 10 (83.3%)
AI impact assessment
AI impact assessment
Reported in progress: 1 (8.3%)AI impact assessment
Reported not in place: 1 (8.3%)AI impact assessment
No answer reported: 10 (83.3%)
12Reported in progress: 1 (8.3%)AI impact assessment
Reported not in place: 1 (8.3%)AI impact assessment
No answer reported: 10 (83.3%)
Independent review12
Ongoing monitoring
Ongoing monitoring
Reported in place: 1 (8.3%)Ongoing monitoring
Other or ambiguous answer: 1 (8.3%)Ongoing monitoring
No answer reported: 10 (83.3%)
12Reported in place: 1 (8.3%)Ongoing monitoring
Other or ambiguous answer: 1 (8.3%)Ongoing monitoring
No answer reported: 10 (83.3%)
Operator training
Operator training
Reported in place: 1 (8.3%)Operator training
Reported in progress: 1 (8.3%)Operator training
No answer reported: 10 (83.3%)
12Reported in place: 1 (8.3%)Operator training
Reported in progress: 1 (8.3%)Operator training
No answer reported: 10 (83.3%)
| 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 | 12 | 2 | 0 | 0 | 0 | 0 | 0 | 10 | 17% | 17% of 12 |
| AI impact assessment | 12 | 0 | 1 | 1 | 0 | 0 | 0 | 10 | 17% | 0% of 12 |
| Independent review | 12 | 0 | 1 | 0 | 0 | 0 | 0 | 11 | 8% | 0% of 12 |
| Ongoing monitoring | 12 | 1 | 0 | 0 | 0 | 0 | 1 | 10 | 17% | 8% of 12 |
| Operator training | 12 | 1 | 1 | 0 | 0 | 0 | 0 | 10 | 17% | 8% of 12 |
| Fail-safe | 12 | 1 | 0 | 0 | 1 | 0 | 0 | 10 | 17% | 9% of 11 |
| Appeal process | 12 | 0 | 0 | 0 | 1 | 0 | 0 | 11 | 8% | 0% of 11 |
| User and public consultation | 12 | 1 | 1 | 0 | 0 | 0 | 0 | 10 | 17% | 8% of 12 |
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 | |
|---|---|---|---|---|---|---|---|---|---|
| EPA | 17% | 0% | 0% | 8% | 8% | 9% | 0% | 8% | 12 |
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 |
|---|---|---|---|---|---|---|---|---|---|
| Streamflow Duration Assessment Modeling EPA · 25 | 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 |
| Microsoft Power Automate (custom document processing model) EPA · 35 | 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 |
| Helpdesk chat assistant for common data system tasks EPA · 39 | 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 |
| Agency Records Management System EPA · 73 | Deployed | Reported in place | Reported not in place | No answer reported | Other or ambiguous answer | Reported in progress | Reported in place | No answer reported | Reported in place |
| Living Literature Review: Semi-automated Literature Screening EPA · 76 | 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 |
| Using machine learning methods to analyze drivers of water quality EPA · 77 | 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 |
| 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 |
| FFEO TSCA investigation into privatized military housing property management companies EPA · 43 | Pre-deployment | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
| OPP Document Data Extraction EPA · 60 | Pre-deployment | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
| CWA PA EPA · 79 | Pre-deployment | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
| CAA PA EPA · 80 | Pre-deployment | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported | No answer reported |
| RMP Risk Modeling EPA · 82 | Pre-deployment | 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- 1 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 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.