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: 38 records · use cases that answered at least one practice question

Reported in placeReported in progressNot applicablePrecluded by law or guidanceOther or ambiguous answerNo answer reported

Pre-deployment testing

Pre-deployment testing
Reported in place: 14 (36.8%)Pre-deployment testing
Reported in progress: 5 (13.2%)Pre-deployment testing
No answer reported: 19 (50.0%)

AI impact assessment

AI impact assessment
Reported in place: 11 (28.9%)AI impact assessment
Reported in progress: 4 (10.5%)AI impact assessment
No answer reported: 23 (60.5%)

Independent review

Independent review
Reported in place: 30 (78.9%)Independent review
Reported in progress: 5 (13.2%)Independent review
No answer reported: 3 (7.9%)

Ongoing monitoring

Ongoing monitoring
Reported in place: 32 (84.2%)Ongoing monitoring
Reported in progress: 5 (13.2%)Ongoing monitoring
No answer reported: 1 (2.6%)

Operator training

Operator training
Reported in place: 33 (86.8%)Operator training
Reported in progress: 5 (13.2%)

Fail-safe

Fail-safe
Reported in place: 10 (26.3%)Fail-safe
Reported in progress: 1 (2.6%)Fail-safe
Not applicable: 27 (71.1%)

Appeal process

Appeal process
Reported in place: 2 (5.3%)Appeal process
Reported in progress: 20 (52.6%)Appeal process
Not applicable: 11 (28.9%)Appeal process
Precluded by law or guidance: 3 (7.9%)Appeal process
No answer reported: 2 (5.3%)

User and public consultation

User and public consultation
Reported in place: 6 (15.8%)User and public consultation
Reported in progress: 9 (23.7%)User and public consultation
Other or ambiguous answer: 9 (23.7%)User and public consultation
No answer reported: 14 (36.8%)

Practice	In scope	In place	In progress	N/A	Precluded or waived	Other	No answer	Answered	In place of applicable
Pre-deployment testing	38	14	5	0	0	0	19	50%	37% of 38
AI impact assessment	38	11	4	0	0	0	23	39%	29% of 38
Independent review	38	30	5	0	0	0	3	92%	79% of 38
Ongoing monitoring	38	32	5	0	0	0	1	97%	84% of 38
Operator training	38	33	5	0	0	0	0	100%	87% of 38
Fail-safe	38	10	1	27	0	0	0	100%	91% of 11
Appeal process	38	2	20	11	3	0	2	95%	8% of 24
User and public consultation	38	6	9	0	0	9	14	63%	16% of 38

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
DOI	–	–	95%	95%	95%	0%	0%	0%	19
DHS	100%	100%	100%	100%	100%	100%	0%	20%	5
SSA	100%	100%	100%	100%	100%	100%	100%	0%	5
SBA	25%	25%	25%	25%	25%	0%	0%	25%	4
DOE	75%	–	25%	75%	100%	100%	0%	100%	4
DOJ	0%	0%	0%	0%	0%	0%	0%	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

27 records with at least one practice unanswered or in progressOpen in explorer

Use case	Stage	Test	AIA	Review	Monitor	Train	Fail-safe	Appeal	Consult
OPQ-AI DOE · DOE-477	Deployed	Reported in place	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	No answer reported	Reported in place
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
pComply-AI-High Risk DOE · DOE-658	Deployed	Reported in place	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	No answer reported	Reported in place
Rangeland Condition Monitoring Assessment and Projection (RCMAP) [2024 INV#WO0000000107126] DOI · DOI-0171	Deployed	No answer reported	No answer reported	Reported in progress	Reported in progress	Reported in progress	Not applicable	Not applicable	Reported in progress
use of random forest for species distribution modeling DOI · DOI-0161	Deployed	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Global food-and-water security-support analysis data (GFSAD) project [2024 INV#WO0000000107073] DOI · DOI-0095	Deployed	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Frog vocalization recognition from digital recordings DOI · DOI-0092	Deployed	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Patterns in the Landscape – Analyses of Cause and Effect DOI · DOI-0090	Deployed	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Submersed Aquatic Vegetation Vulnerability Evaluation Application (SAVVEA) DOI · DOI-0081	Deployed	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Machine learning in remote sensing-based wildfire and natural resource risk assessments DOI · DOI-0072	Deployed	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Telemetry Analysis Learning Algorithm (TALA) DOI · DOI-0070	Deployed	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
LANDFIRE DOI · DOI-0069	Deployed	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Static-99 Sex Offender Data System (SODS) DOJ · DOJ-0163	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
Crickets DOE · DOE-723	Pre-deployment	Reported in progress	No answer reported	No answer reported	No answer reported	Reported in place	Not applicable	Not applicable	Reported in place
Deep Learning based image segmentation [2024 INV#WO0000000107975] DOI · DOI-0175	Pilot	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Data-driven approaches to filling missing time-series data within the San Francisco Bay-Delta [2024 INV#WO0000000107683; INV#WO0000000107488] DOI · DOI-0174	Pre-deployment	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Machine Learning based shoreline detection and sea ice dynamics using coastal cameras [2024 INV#WO0000000108008] DOI · DOI-0166	Pre-deployment	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	Reported in progress
Tsunami Hazard Analysis [2024 INV#WO0000000108470] DOI · DOI-0165	Pre-deployment	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	Reported in progress
Determining the resource potential of critical minerals in seafloor massive sulfide deposits [2024 INV#WO0000000109311] DOI · DOI-0163	Pre-deployment	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	Reported in progress
Reducing elevation error in coastal wetland digital elevation models DOI · DOI-0088	Pilot	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Developing land cover maps for barrier islands using satellite imagery DOI · DOI-0087	Pilot	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Invasive Carp Harvest Predictive Model DOI · DOI-0061	Pilot	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Automated identification of ducks from hunter-submitted photos using deep learning models DOI · DOI-0074	Pilot	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	No answer reported
Pathogen identification in salmon DOI · DOI-0071	Pilot	No answer reported	No answer reported	Reported in place	Reported in place	Reported in place	Not applicable	Reported in progress	Reported in progress
Audit Green Book Classification SBA · SBA-3	Pre-deployment	Reported in progress	Reported in progress	Reported in progress	Reported in progress	Reported in progress	Not applicable	Not applicable	Reported in progress
Internal OFO Chatbot Recommender SBA · SBA-4	Pre-deployment	Reported in progress	Reported in progress	Reported in progress	Reported in progress	Reported in progress	Not applicable	Not applicable	Reported in progress
7(a) Purchases ML Model SBA · SBA-7	Pre-deployment	Reported in progress	Reported in progress	Reported in progress	Reported in progress	Reported in progress	Not applicable	Reported in progress	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?
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?
5 high-impact use cases mark an appeal process as not applicable. On what basis was each individual outcome judged not contestable?
6 high-impact use cases report using PII but give no link to a Privacy Impact Assessment. Is a PIA published elsewhere?
4 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.