Technology landscape
How reported AI classifications spread across agencies, missions, lifecycle stages and sourcing. Each record carries one reported classification, so categories do not overlap. A classification describes the technique, not a specific commercial model.
Lifecycle by technology
Share of each technology's recordsSourcing by technology
Where reportedDeveloped in-house: 37 (14.7%)Generative AI
Contract and in-house: 76 (30.2%)Generative AI
Purchased from vendor: 130 (51.6%)Generative AI
Not reported: 9 (3.6%)
Developed in-house: 149 (40.7%)Classical / predictive ML
Contract and in-house: 83 (22.7%)Classical / predictive ML
Purchased from vendor: 129 (35.2%)Classical / predictive ML
Not reported: 5 (1.4%)
Developed in-house: 60 (29.9%)Natural language processing
Contract and in-house: 69 (34.3%)Natural language processing
Purchased from vendor: 66 (32.8%)Natural language processing
Not reported: 6 (3.0%)
Developed in-house: 12 (7.4%)Computer vision
Contract and in-house: 28 (17.3%)Computer vision
Purchased from vendor: 120 (74.1%)Computer vision
Not reported: 2 (1.2%)
Developed in-house: 8 (40.0%)Agentic AI
Contract and in-house: 7 (35.0%)Agentic AI
Purchased from vendor: 5 (25.0%)
Developed in-house: 2 (40.0%)Reinforcement learning
Contract and in-house: 2 (40.0%)Reinforcement learning
Purchased from vendor: 1 (20.0%)
Technology adoption by agency
Row share: each agency's records by classification · 16 largest portfolios in the filter| Generative AI | Classical / predictive ML | Natural language processing | Computer vision | Agentic AI | Reinforcement learning | Other | Not reported | n | |
|---|---|---|---|---|---|---|---|---|---|
| HHS | 26% | 28% | 33% | 8% | 4% | 2% | 167 | ||
| DOJ | 18% | 44% | 21% | 15% | 1% | 163 | |||
| VA | 6% | 26% | 17% | 49% | 2% | 138 | |||
| DOE | 57% | 18% | 8% | 2% | 2% | 3% | 10% | 119 | |
| DHS | 19% | 37% | 12% | 32% | 110 | ||||
| DOI | 78% | 13% | 5% | 3% | 3% | 40 | |||
| USDA | 21% | 54% | 15% | 8% | 3% | 39 | |||
| SSA | 33% | 44% | 7% | 11% | 4% | 27 | |||
| NASA | 50% | 15% | 19% | 15% | 26 | ||||
| TREAS | 31% | 35% | 35% | 26 | |||||
| DOL | 33% | 57% | 5% | 5% | 21 | ||||
| SEC | 35% | 35% | 10% | 15% | 5% | 20 | |||
| STATE | 47% | 16% | 26% | 11% | 19 | ||||
| GSA | 35% | 12% | 35% | 12% | 6% | 17 | |||
| FRB | 6% | 31% | 63% | 16 | |||||
| FDIC | 6% | 25% | 44% | 13% | 13% | 16 |
Technology by topic area
Row share within each topic| Generative AI | Classical / predictive ML | Natural language processing | Computer vision | Agentic AI | Reinforcement learning | Other | Not reported | n | |
|---|---|---|---|---|---|---|---|---|---|
| Law Enforcement | 12% | 37% | 18% | 30% | 2% | 1% | 179 | ||
| Administrative Functions | 37% | 29% | 22% | 7% | 2% | 1% | 2% | 171 | |
| Health and Medical | 6% | 30% | 21% | 42% | 1% | 155 | |||
| Information Technology | 58% | 23% | 11% | 1% | 2% | 5% | 124 | ||
| Other | 23% | 33% | 26% | 4% | 1% | 1% | 12% | 100 | |
| Science | 6% | 64% | 8% | 15% | 6% | 1% | 87 | ||
| Procurement and Financial Management | 29% | 41% | 22% | 2% | 5% | 41 | |||
| Service Delivery | 22% | 29% | 44% | 5% | 41 | ||||
| Energy and the Environment | 38% | 38% | 8% | 8% | 8% | 26 | |||
| Government Benefits Processing | 12% | 46% | 27% | 12% | 4% | 26 | |||
| Human Resources | 56% | 20% | 12% | 8% | 4% | 25 | |||
| Transportation | 13% | 22% | 26% | 9% | 4% | 26% | 23 | ||
| Cybersecurity | 27% | 64% | 5% | 5% | 22 | ||||
| Emergency Management | 64% | 27% | 9% | 11 |
High-impact designation by technology
High-impact: 15 (6.0%)Generative AI
Presumed high-impact, determined not: 8 (3.2%)Generative AI
Not high-impact: 221 (87.7%)Generative AI
Not reported: 8 (3.2%)
High-impact: 86 (23.5%)Classical / predictive ML
Presumed high-impact, determined not: 33 (9.0%)Classical / predictive ML
Not high-impact: 245 (66.9%)Classical / predictive ML
Not reported: 2 (0.5%)
High-impact: 26 (12.9%)Natural language processing
Presumed high-impact, determined not: 5 (2.5%)Natural language processing
Not high-impact: 164 (81.6%)Natural language processing
Not reported: 6 (3.0%)
High-impact: 98 (60.5%)Computer vision
Presumed high-impact, determined not: 14 (8.6%)Computer vision
Not high-impact: 48 (29.6%)Computer vision
Not reported: 2 (1.2%)
Governance disclosure by technology
High-impact records: reported in place / scope| Technology | High-impact | Testing | Impact assess. | Indep. review | Monitoring | Training |
|---|---|---|---|---|---|---|
| Generative AI | 15 | 2 / 15 | 3 / 15 | 3 / 15 | 2 / 15 | 4 / 15 |
| Classical / predictive ML | 86 | 19 / 86 | 16 / 86 | 16 / 86 | 19 / 86 | 18 / 86 |
| Natural language processing | 26 | 4 / 26 | 2 / 26 | 2 / 26 | 2 / 26 | 2 / 26 |
| Computer vision | 98 | 19 / 98 | 15 / 98 | 15 / 98 | 17 / 98 | 17 / 98 |
| Agentic AI | 2 | 0 / 2 | 0 / 2 | 0 / 2 | 0 / 2 | 0 / 2 |
| Reinforcement learning | 0 | – | – | – | – | – |
| Other | 0 | – | – | – | – | – |
| Not reported | 0 | – | – | – | – | – |
Shows counts reported in place against all high-impact records of that technology. The remainder includes in-progress, not applicable and unanswered. Full governance view
Systems, vendors and technology are different things
551 of 1,040 records name a vendor (53%)Vendor: the supplier the agency names in a free-text field. Matched to a standard list of companies; one record may name several.
System name: the agency's name for the system (reported for 63% of records).
AI classification: the technique category the agency selected. A vendor name does not show which model or technique is used, and this platform does not infer it.