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.

Governance report

Practice-by-practice evidence

Denominator: 31 records · all use cases in the current filter

Reported 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: 31 (100.0%)

AI impact assessment

AI impact assessment
Reported in place: 19 (61.3%)AI impact assessment
Reported in progress: 5 (16.1%)AI impact assessment
No answer reported: 7 (22.6%)

Independent review

Independent review
Reported in place: 26 (83.9%)Independent review
Reported in progress: 2 (6.5%)Independent review
Waived by CAIO: 1 (3.2%)Independent review
No answer reported: 2 (6.5%)

Ongoing monitoring

Ongoing monitoring
Reported in place: 26 (83.9%)Ongoing monitoring
Reported in progress: 3 (9.7%)Ongoing monitoring
No answer reported: 2 (6.5%)

Operator training

Operator training
Reported in place: 27 (87.1%)Operator training
Reported in progress: 2 (6.5%)Operator training
Other or ambiguous answer: 1 (3.2%)Operator training
No answer reported: 1 (3.2%)

Fail-safe

Fail-safe
Not applicable: 31 (100.0%)

Appeal process

Appeal process
Reported in place: 2 (6.5%)Appeal process
Not applicable: 28 (90.3%)Appeal process
Precluded by law or guidance: 1 (3.2%)

User and public consultation

User and public consultation
Reported in place: 21 (67.7%)User and public consultation
Reported in progress: 1 (3.2%)User and public consultation
Other or ambiguous answer: 9 (29.0%)

Practice	In scope	In place	In progress	N/A	Precluded or waived	Other	No answer	Answered	In place of applicable
Pre-deployment testing	31	31	0	0	0	0	0	100%	100% of 31
AI impact assessment	31	19	5	0	0	0	7	77%	61% of 31
Independent review	31	26	2	0	1	0	2	94%	87% of 30
Ongoing monitoring	31	26	3	0	0	0	2	94%	84% of 31
Operator training	31	27	2	0	0	1	1	97%	87% of 31
Fail-safe	31	0	0	31	0	0	0	100%	n/a of 0
Appeal process	31	2	0	28	1	0	0	100%	100% of 2
User and public consultation	31	21	1	0	0	9	0	100%	68% of 31

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.

Agency-by-safeguard heatmap

Cell: share reported in place, among applicable recordsIn place Answered

Agency by governance practice
	Testing	Impact assess.	Indep. review	Monitoring	Training	Fail-safe	Appeal	Consultation	n
SBA	100%	100%	100%	100%	100%	0%	0%	82%	11
DOE	100%	9%	80%	64%	73%	0%	0%	91%	11
DHS	100%	83%	83%	83%	100%	0%	0%	33%	6
DOL	100%	100%	100%	100%	0%	0%	100%	0%	1
EPA	100%	0%	0%	100%	100%	0%	0%	0%	1
SSA	100%	100%	100%	100%	100%	0%	100%	0%	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
CBP Translate DHS · DHS-2388	Deployed	Reported in place	Reported in progress	Reported in progress	Reported in progress	Reported in place	Not applicable	Not applicable	Reported in place
ServiceNow Predictive Intelligence DOE · DOE-430	Deployed	Reported in place	Reported in progress	Reported in place	Reported in place	Reported in place	Not applicable	Not applicable	Reported in place
Hanford Ai Liaison (HAL) 1.1 DOE · DOE-469	Deployed	Reported in place	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Not applicable	Reported in place
LANL AI Portal DOE · DOE-495	Deployed	Reported in place	Reported in progress	Reported in place	Reported in place	Reported in place	Not applicable	Not applicable	Reported in place
ChatGPT Enterprise DOE · DOE-498	Deployed	Reported in place	Reported in progress	Reported in place	Reported in place	Reported in place	Not applicable	Not applicable	Reported in place
GitHub Co-Pilot DOE · DOE-571	Deployed	Reported in place	No answer reported	Reported in place	Reported in progress	Reported in progress	Not applicable	Not applicable	Reported in place
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
Invoice Scanning System DOE · DOE-632	Deployed	Reported in place	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Not applicable	Reported in place
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
Microsoft Co-Pilot DOE · DOE-518	Pilot	Reported in place	No answer reported	Reported in place	Reported in progress	Reported in progress	Not applicable	Not applicable	Reported in place
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
Hanford Ai Liaison (HAL) 2.0 DOE · DOE-653	Pre-deployment	Reported in place	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Not applicable	Reported in place

Unanswered analytical questions

Generated from the data; for follow-up with agencies

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?
6 high-impact use cases mark an appeal process as not applicable. On what basis was each individual outcome judged not contestable?
2 high-impact use cases report using PII but give no link to a Privacy Impact Assessment. Is a PIA published elsewhere?
14 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.