AI-Tagged Federal Contracts Jump 61% Year-Over-Year
Summary
Federal contracts whose USAspending record matches "artificial intelligence" drew $121.3 million in obligations across 81 awards in the January–March 2026 quarter — up 61% from $75.5 million a year earlier, and more than seven times the $14.7 million obligated the prior quarter. The Department of Defense wrote 83% of it, with the Army alone at $50.2 million; civilian agencies, about 5% of the total a year ago, supplied a growing share of the rest.
View data as table
| FY2025 Q1 | $60.3M | Oct–Dec 2024 |
|---|---|---|
| FY2025 Q2 | $75.5M | Jan–Mar 2025 (YoY base) |
| FY2025 Q3 | $65.7M | Apr–Jun 2025 |
| FY2025 Q4 | $184.8M | Jul–Sep 2025, window peak |
| FY2026 Q1 | $14.7M | Oct–Dec 2025, window low |
| FY2026 Q2 | $121.3M | Jan–Mar 2026, 81 awards |
The rebound is real but not a record: the same segment obligated $184.8 million in the July–September 2025 quarter before collapsing to a window-low $14.7 million over the winter. Read against a year earlier, though, the direction is unambiguous — 81 AI-tagged awards this quarter versus 55, and $45.8 million more money.
Defense wrote 83 cents of every dollar
The transaction-level records⧉ that make up the quarter — 158 obligations that sum exactly to the $121,276,464.47 headline — are concentrated hard in Defense: $100.5 million, with the Army alone at $50.2 million and the Missile Defense Agency and Air Force close behind. One vendor, ECS Federal, LLC, took $23.2 million across several Defense actions. The broadening happened on the civilian side: Health and Human Services ($7.7M), Commerce ($6.1M) and Treasury ($5.5M) each posted AI-tagged awards this quarter after DoD supplied roughly 95% of the far smaller total a year before.
View data as table
| Department of Defense | $100.5M | 82.8% of total |
|---|---|---|
| Health & Human Services | $7.7M | 6.3% |
| Commerce | $6.1M | 5.0% |
| Treasury | $5.5M | 4.5% |
| All other agencies (net) | $1.6M | 1.3% |
- $121.3M across 81 AI-tagged federal contract awards in Q2 FY2026 (Jan–Mar 2026) — up 61% from $75.5M and 55 awards a year earlier.
- More than 7x the prior quarter's $14.7M — the segment's largest quarter-over-quarter increase in the eight quarters on record, though still below FY2025 Q4's $184.8M.
- Defense drove 83% of it ($100.5M), the Army the largest sub-agency at $50.2M and ECS Federal the top vendor at $23.2M.
- Civilian agencies broadened the base: , Commerce and Treasury each posted AI-tagged awards, versus a roughly 5% civilian share ($3.6M) a year earlier.
"AI-tagged" means the award record matches the keyword "artificial intelligence" in USAspending's search (award description, recipient and related text); it is a proxy for AI-related spending, not an official program category, and will both miss AI work described in other words and catch contracts where the phrase is incidental. Figures are federal-action obligations on definitive contract types A–D, counted by transaction action date, so a large multi-year award (such as ECS Federal's $120.6M R&D vehicle first signed in FY2020) counts only the dollars obligated within the quarter, not its lifetime value. USAspending is a live database; recent-quarter totals revise as agencies report, so these are a point-in-time snapshot pulled 2026-08-14. The full pulled dataset, reconciling the 158 transactions to the $121,276,464.47 aggregate, ships beside this article as data-usaspending-fy2026q2.json.
Sources(5) ▾
- U.S. Department of the Treasury / USAspending.gov (federal award data), USAspending.gov — Spending Over Time (quarter), AI-keyword contracts, FY2025 Q1 – FY2026 Q4 (2026-08-14) — Reproducible POST to the official federal spending API. Body: {"group":"quarter","filters":{"award_type_codes":["A","B","C","D"],"keywords":["artificial intelligence"],"time_period":[{"start_date":"2024-10-01","end_date":"2026-08-14"}]}}. Returns Contract_Obligations aggregated by federal fiscal quarter on transactions whose award record matches the keyword "artificial intelligence" (matched against award description, recipient, and related text) for definitive contract award types A/B/C/D. This is the load-bearing series for every quarterly figure in the piece. Pulled 2026-08-14; the full response ships beside this article as data-usaspending-fy2026q2.json (field quarterly_obligations_over_time). USAspending is a live database — obligations for recent quarters revise as agencies report; values are a point-in-time snapshot. api.usaspending.gov · original document
- U.S. Department of the Treasury / USAspending.gov (federal award data), USAspending.gov — Spending by Award Count, AI-keyword contracts, Jan–Mar 2026 (Q2 FY2026) (2026-08-14) — Reproducible POST. Body: {"filters":{"award_type_codes":["A","B","C","D"],"keywords":["artificial intelligence"],"time_period":[{"start_date":"2026-01-01","end_date":"2026-03-31","date_type":"action_date"}]}}. Returns results.contracts = the count of distinct prime contract awards with a matching transaction whose action_date falls in the window. Pulled 2026-08-14; response value stored in data-usaspending-fy2026q2.json (award_counts.Q2FY2026_Jan-Mar-2026). api.usaspending.gov · original document
- U.S. Department of the Treasury / USAspending.gov (federal award data), USAspending.gov — Spending by Award Count, AI-keyword contracts, Jan–Mar 2025 (Q2 FY2025) (2026-08-14) — Reproducible POST, identical to doc-usa-count-q2fy26 except time_period start_date 2025-01-01, end_date 2025-03-31. Returns results.contracts for the prior-year quarter, the denominator for the year-over-year award-count comparison. Pulled 2026-08-14; stored in data-usaspending-fy2026q2.json (award_counts.Q2FY2025_Jan-Mar-2025). api.usaspending.gov · original document
- U.S. Department of the Treasury / USAspending.gov (federal award data), USAspending.gov — Spending by Transaction, AI-keyword contracts, Jan–Mar 2026 (Q2 FY2026) (2026-08-14) — Reproducible POST (paginated, limit 100). Body: {"filters":{"award_type_codes":["A","B","C","D"],"keywords":["artificial intelligence"],"time_period":[{"start_date":"2026-01-01","end_date":"2026-03-31","date_type":"action_date"}]},"fields":["Award ID","Recipient Name","Transaction Amount","Action Date","Awarding Agency","Awarding Sub Agency"],"sort":"Transaction Amount","order":"desc"}. Returns the individual transactions that make up the quarter: 158 transactions across the 81 awards, whose Transaction Amount values sum to $121,276,464.47 — reconciling exactly to the Spending Over Time aggregate for FY2026 Q2. Grouping those transaction amounts by Awarding Agency and Recipient Name yields the agency and vendor concentration reported here. Full agency, sub-agency, and top-recipient rollups plus the twelve largest transactions ship in data-usaspending-fy2026q2.json (Q2FY2026 block). IMPORTANT distinction: a prime award's lifetime 'Award Amount' (e.g. the $120.6M ECS Federal R&D vehicle first signed in FY2020) is NOT its FY2026 Q2 obligation; only the transaction obligations dated in the window count toward the quarter, which is why this transaction-level pull, not the award-level totals, is used to attribute the trend. api.usaspending.gov · original document
- U.S. Department of the Treasury / USAspending.gov (federal award data), USAspending.gov — Spending by Transaction, AI-keyword contracts, Jan–Mar 2025 (Q2 FY2025) (2026-08-14) — Reproducible POST, identical to doc-usa-txn-q2fy26 except time_period start_date 2025-01-01, end_date 2025-03-31. Returns the prior-year quarter's 134 transactions summing to $75,504,359.15 (reconciling to Spending Over Time FY2025 Q2); grouped by Awarding Agency, the Department of Defense accounts for $71,873,731.18 of that total. Used for the year-over-year composition comparison. Stored in data-usaspending-fy2026q2.json (Q2FY2025 block). api.usaspending.gov · original document
Comments
Always open. Logged-in readers can annotate paragraphs in place.
Federal definitive contracts whose USAspending.gov⧉ record matches the keyword "artificial intelligence" drew $121.3 million in obligations across 81 awards in fiscal-year 2026's second quarter (January–March 2026) — up 61% from $75.5 million in the same quarter a year earlier, and more than seven times the $14.7 million obligated the quarter before. It is the largest quarter-over-quarter increase in this AI-tagged segment across the eight quarters on record.