{"resourceId":"doj-oig-nij-purdue-ai-community-supervision-26-051","versions":[{"version":"external-98e92b35ef28123ecfefdb3058d09defcda96a5a59e4b98ad9ba05be7b1af096","resource":{"id":"doj-oig-nij-purdue-ai-community-supervision-26-051","title":"DOJ OIG 26-051: NIJ Purdue AI community-supervision award spent ~95% without completing goals","organization":"U.S. Department of Justice Office of the Inspector General","sector":"Federal oversight of corrections / community supervision R&D","geography":"United States (Purdue University, West Lafayette, IN; multi-subrecipient)","publishedAt":"May 2026 (posted May 7, 2026)","publicationDate":"2026-05-07","eventDate":null,"sourceName":"DOJ Office of the Inspector General","sourceLabel":"Federal government audit","sourceUrl":"https://oig.justice.gov/sites/default/files/reports/26-051.pdf","evidenceClass":"government-audit","outcomeClass":"cautionary","topics":["infrastructure","data-security","governance-procurement","operating-model","accessibility-workforce"],"finding":"DOJ OIG Audit 26-051 (May 2026) reviewed NIJ’s $1,999,778 cooperative agreement to Purdue for an AI intervention system to support parolee reentry (smartphones/smartwatches to identify positive activities, alert risky behaviors, and inform supervision). OIG found Purdue developed a system but could not fully deploy/analyze efficacy; spent $1,908,515 (~95%) without fully completing goals due largely to human-subjects non-compliance and weak NIJ/Purdue oversight; 8 of 35 parolee participants absconded with federally purchased devices without full awareness/recovery; questioned costs $86,036 and $91,263 funds for better use; 11 recommendations on oversight, risk controls for tech to offenders, subrecipient monitoring, and performance documentation. Why now: not previously archived; newly salient beside 19 Sep Tasmania parole AI governance failures—shows federal AI community-supervision R&D can consume nearly all funds without validated outcomes when risk controls lag.","sledRelevance":"SLED corrections/community-supervision buyers and state justice grant offices should treat AI reentry wearables/apps as high-oversight procurements: human-subjects compliance, absconder/device recovery plans, and independent outcome evaluation—not lab demos alone.","evidence":"Full OIG audit PDF with executive summary metrics (spend, absconders 8/35, questioned costs, 11 recommendations). Covers award Jan 2020–Dec 2023 performance window; posted May 2026. Does not evaluate contemporary commercial supervision AI products beyond this award.","architectureImplications":"Interpretation: supervision AI that issues devices to clients needs inventory, remote wipe/recovery, and alerting when devices leave control; evaluation instrumentation must survive stop-work and IRB constraints.","governanceImplications":"Interpretation: cooperative agreements require substantial federal involvement and documented risk frameworks before tech touches supervised persons; subrecipient monitoring is non-optional.","securityPrivacyImplications":"Federally purchased phones/watches on parolees create safety/security exposure when absconded; research protocols must address loss, misuse, and data on devices.","caveats":"May 2026 publication (older than last run) included for new archive coverage and relevance to current corrections AI governance attention; single award; redactions in grantee response appendix; not a product bake-off.","streamIds":["public-safety"],"roles":{"sales":"Sales framing: OIG shows a nearly fully spent AI community-supervision R&D award that failed to finish deployment/evaluation and lost track of absconder devices. Sell oversight design, device lifecycle controls, and evaluation gates—not unverified recidivism claims. Stakeholders: DOC/community supervision directors, state administering agencies, IRB/compliance, finance. Bounded engagement: pre-award risk framework and device-recovery plan review. Unsupported: that commercial apps avoid these failure modes.","engineering":"Engineering framing: Fit only with inventory/MDM, absconder playbooks, and outcome measurement wired before enrollment. Prerequisites: IRB clearance windows, stop-work compliance, subrecipient controls. Constraints: human-subjects pauses can strand spend without outcomes. Security: recover/wipe lost devices; minimize sensitive data at rest on wearables. Proof of value: completed deployment+outcome analysis under an approved protocol—not drawdown percentage.","delivery":"Delivery framing: Stage build→controlled pilot→outcome analysis with explicit go/no-go; monitor parole status daily for device recovery. Dependencies: NIJ/state grant conditions, IRB, supervision officers, MDM vendor. Ownership: PI for research integrity; supervision agency for client/device custody; grant office for fiscal allowability. Skills: research compliance literacy for technologists. Acceptance: goals met or funds stopped; zero untracked devices; questioned-cost controls. Risks: absconder safety incidents; questioned costs; incomplete evaluations after large spend."},"roleDetails":{"sales":{"customerProblem":"Agencies and grantors fund AI supervision tools without assurance of deployment, evaluation, or device control when clients abscond.","stakeholders":["Community supervision director","State administering agency / NIJ grantee office","IRB/compliance","Finance/grants","Field officers"],"discoveryQuestions":["What is the device recovery plan if a client absconds?","Which outcome metrics gate continued funding?","How are stop-work/IRB pauses handled financially?","Who monitors subrecipients daily?"],"valueHypothesis":"Bounded value is supervised pilot with recoverable devices and completed evaluation—not 95% burn rate.","potentialEngagement":"Pre-procurement risk and evaluation design review against OIG 26-051 control themes.","unsupportedClaims":["Recidivism reduction from this Purdue system","That questioned costs are unique to academia","Guaranteed absconder recovery"]},"engineering":{"fit":"AI sensing/alerting for community supervision only inside a controlled research or ops envelope with MDM.","architectureIntegration":"Device MDM + event telemetry → officer workflow → research warehouse; kill-switch and inventory reconciliation jobs.","prerequisites":["IRB protocol covering device loss","MDM/remote wipe","Subrecipient monitoring plan","Performance metric store"],"constraints":["Human-subjects stop-work risk","Multi-subrecipient coordination","Historical award context"],"security":"Encrypt devices; rapid wipe on abscond alert; bar unnecessary PII sync to personal clouds.","proofOfValue":["Deployment completion rate","Outcome analysis delivered","Device recovery SLA met"]},"delivery":{"work":"Implement enrollment controls, daily parole-status/device checks, fiscal allowability reviews, and evaluation milestones.","dependencies":["Supervision case data","MDM","IRB","Grant special conditions"],"ownership":"Agency supervision lead owns clients/devices; PI owns analysis; grants office owns allowability.","skillsAdoption":"Train officers on device custody and research boundary rules.","governanceCheckpoints":["IRB clearance","Pilot enrollment cap review","Mid-award outcome gate","Closeout device audit"],"acceptanceCriteria":["Evaluation report completed","All devices accounted or documented lost with wipe attempt","No unapproved human-subjects work"],"risks":["Absconder incidents with federal devices","Questioned costs","Spend without outcomes"]}},"retrievedAt":"2026-09-21T03:08:50Z","enrichedAt":"2026-09-21T03:22:00Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Field officers need clear procedures for device issuance/recovery; research staff need compliance training beyond model building.","procurementImplications":"Grant RFPs should require device lifecycle plans, subrecipient monitoring, and evaluation deliverables tied to payment milestones.","operatingModelImplications":"Separate research build from operational custody of client devices; escalate absconder events to both research and supervision chains.","sourceVerification":{"openedUrl":"https://oig.justice.gov/sites/default/files/reports/26-051.pdf","referenceExcerpt":"Purdue expended $1,908,515, or 95 percent of the cooperative agreement funding, but did not fully complete award activities","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}