LO
DeploymentProcess & self-service2019-09-01
The US Department of Justice reports a department-wide system, operational since September 2019, that triages and classifies IT helpdesk tickets and handles common requests with virtual agents
IT support specialist / helpdeskoccupation page →
Event date / reported
2019-09-01 · reported 2026-04-14
Evidence stage
DeploymentAn employer has put it into production. Can move the baseline — weighted by scale and how similar the setting is.
Tasks this bears on
The ticket you have seen four hundred times
Password resets, account unlocks, access requests, the printer, the VPN — the same twenty problems that make up most of the queue.
Automating✓ Evidence-backed
Where this applies
The entry is the department's own submission to the statutory inventory OMB publishes under the Advancing American AI Act, so the publisher here is the employer describing a deployment it runs, not a vendor or a journalist. The entry gives its operational date only as a month, 09/2019, so it is recorded as the first of that month; the reported date is the April 2026 data upload, which the repository README dates as of 13 April 2026. Read across the whole 2025 file and hand-filtered to genuine IT service-desk use cases, the same inventory holds 12 deployed across 6 agencies, 4 pilots, 26 in pre-deployment across 11 agencies and 1 retired — more are waiting than are running. What the entry does not give is any headcount figure, and it records no measurement of what share of the queue the system closes. Because this deployment predates public generative models by three years, it is evidence about ticket triage automation, not about language models.
What this means
One very large employer has been running ticket triage, classification and a virtual agent for common requests since 2019, and says so in a report it is legally obliged to file. For a service desk, that dates the automation of the narrow, repetitive tier to before generative models existed, which makes the useful question not whether it arrives but which tier of the queue your day is actually made of.
What it does not yet show
It does not establish that anyone's headcount changed. The inventory carries no staffing figure and no measure of what share of the queue the system closes, so the size of the effect is unknown from this source. It also does not show a uniform picture: in the same file 26 service-desk use cases across 11 agencies are recorded as not yet deployed, against 12 deployed across 6.
What you can check
Open your own ticketing tool's report for last month and find the share of tickets closed without a human opening them. Most tools report it as tier-0 or self-service resolution. That one number tells you which tier of your queue is already gone, and it is yours rather than a vendor's.
Does it change the assessment?
No. The impact index is never moved by a single event. What this record did: the 1 linked task judgement above now rest on evidence instead of inference.
Source
OMB — 2025 Federal Agency AI Use Case Inventory (entry DOJ-0107) · verified 2026-09-21 · Claude (VOLO agent) — the 3.1 MB individually-reported CSV downloaded from the repository and parsed locally; entry DOJ-0107's own problem_solved, benefits, system_outputs, operational_date (09/2019) and Department-Wide bureau read verbatim, and the 12/4/26/1 stage counts computed over the whole 3,611-row file after hand-filtering out three keyword false positives (elk GPS collars, DEA drones, stroke CT perfusion) that matched on 'support' · interpreted 2026-09-21 · Claude (VOLO agent)
Primary source — published by the party that did this, or the authority of record. No co-signature needed.