{"resourceId":"texas-dir-ai-operating-framework","versions":[{"version":"legacy/2026-08-29/texas-dir-ai-operating-framework","resource":{"id":"texas-dir-ai-operating-framework","title":"Texas turns AI legislation into shared governance and enablement services","organization":"Texas Department of Information Resources","sector":"State and local government technology governance","geography":"Texas, United States","publishedAt":"August 14, 2026","sourceName":"AI in Texas: DIR Implementation of Laws from the 89th Legislature","sourceLabel":"Texas DIR implementation update","sourceUrl":"https://dir.texas.gov/news/ai-texas-dir-implementation-laws-89th-legislature","evidenceClass":"government-evaluation","outcomeClass":"emerging","topics":["knowledge-work","developers-agents","infrastructure","data-security","governance-procurement","accessibility-workforce","operating-model"],"finding":"Texas DIR reports implementing a legislative AI framework through a dedicated AI Division, government AI inventories, a code of ethics and heightened-scrutiny rules, a public-sector sandbox, model policy, certified awareness training, literacy programs, evaluation support, and cooperative contracts.","sledRelevance":"This is a concrete state-level operating model for translating legislation into reusable capabilities that state agencies, local governments, colleges, and school districts can consume rather than interpret independently.","evidence":"DIR reports 96 certified awareness-training programs, more than 27 AI cooperative contracts, six AI Days events, 60 customer lab visits, 150 participating vendors, 40 engaged public-sector organizations, and 76 technology areas showcased. These are implementation and reach metrics, not outcome measures.","architectureImplications":"A shared SLED AI layer can combine sandboxing, evaluation, model and vendor access, policy templates, inventories, training, and data or security guidance while agencies retain ownership of use-case decisions and production controls.","governanceImplications":"Connect statutory duties to an operating catalog: inventory, risk tiering, acceptable use, disclosure, training, sandbox entry and exit criteria, procurement vehicles, evaluation evidence, and accountable local owners.","securityPrivacyImplications":"Heightened-scrutiny rules and sandbox services should be paired with data classification, identity, logging, model isolation, adversarial testing, incident response, and transition criteria before a pilot reaches production.","caveats":"DIR's update is self-reported government implementation evidence. Participation counts do not demonstrate safer systems, improved services, workforce productivity, or public value, and the long-term effect of the framework remains unmeasured."}},{"version":"enrichment/2026-09-05T02:33:27.019Z/texas-dir-ai-operating-framework","resource":{"id":"texas-dir-ai-operating-framework","title":"Texas turns AI legislation into shared governance and enablement services","organization":"Texas Department of Information Resources","sector":"State and local government technology governance","geography":"Texas, United States","publishedAt":"August 14, 2026","publicationDate":"2026-08-14","eventDate":null,"sourceName":"AI in Texas: DIR Implementation of Laws from the 89th Legislature","sourceLabel":"Texas DIR implementation update","sourceUrl":"https://dir.texas.gov/news/ai-texas-dir-implementation-laws-89th-legislature","evidenceClass":"government-evaluation","outcomeClass":"emerging","topics":["knowledge-work","developers-agents","infrastructure","data-security","governance-procurement","accessibility-workforce","operating-model"],"finding":"Texas DIR reports implementing a legislative AI framework through a dedicated AI Division, government AI inventories, a code of ethics and heightened-scrutiny rules, a public-sector sandbox, model policy, certified awareness training, literacy programs, evaluation support, and cooperative contracts.","sledRelevance":"This is a concrete state-level operating model for translating legislation into reusable capabilities that state agencies, local governments, colleges, and school districts can consume rather than interpret independently.","evidence":"DIR reports 96 certified awareness-training programs, more than 27 AI cooperative contracts, six AI Days events, 60 customer lab visits, 150 participating vendors, 40 engaged public-sector organizations, and 76 technology areas showcased. These are implementation and reach metrics, not outcome measures.","architectureImplications":"A shared SLED AI layer can combine sandboxing, evaluation, model and vendor access, policy templates, inventories, training, and data or security guidance while agencies retain ownership of use-case decisions and production controls.","governanceImplications":"Connect statutory duties to an operating catalog: inventory, risk tiering, acceptable use, disclosure, training, sandbox entry and exit criteria, procurement vehicles, evaluation evidence, and accountable local owners.","securityPrivacyImplications":"Heightened-scrutiny rules and sandbox services should be paired with data classification, identity, logging, model isolation, adversarial testing, incident response, and transition criteria before a pilot reaches production.","caveats":"DIR's update is self-reported government implementation evidence. Participation counts do not demonstrate safer systems, improved services, workforce productivity, or public value, and the long-term effect of the framework remains unmeasured.","streamIds":["state-government","local-government","student-success","research","campus-operations","k12"],"roles":{"sales":"Interpretation — Customer problem: agencies and smaller public bodies may lack capacity to translate AI policy into training, evaluation, and usable procurement paths. Stakeholders: statewide technology leadership, agency/local program owners, education IT, procurement, legal, security, and workforce teams. Discovery: which shared services are available and eligible; where do local approval duties remain; and what prevents a sandbox pilot from reaching controlled production? Value hypothesis: reusable governance and enablement services may reduce duplicated effort while preserving local ownership. Potential engagement: map a body's needs to available shared capabilities and plan one use-case evaluation. Unsupported claims: DIR's participation, contract, and training counts are self-reported reach metrics, not demonstrated safety, productivity, or public value; other jurisdictions need their own legal mapping.","engineering":"Interpretation — Fit: evaluate a shared AI enablement layer where agencies can use common sandbox, evaluation, model/vendor access, inventory, and training services. Architecture and integration: connect those services to local identity, data classification, logging, and production approval rather than assuming sandbox access authorizes deployment. Prerequisites: eligibility, service documentation, use-case ownership, applicable rules, and a defined transition path. Constraints: each agency's data, integration, and risk needs may exceed the shared offering, and reported program reach does not demonstrate technical performance. Security: validate model isolation, adversarial testing, access controls, and incident responsibilities at the shared/local boundary. Proposed proof: take one representative use case through sandbox entry, evaluation, and production readiness, documenting unmet requirements and handoff decisions.","delivery":"Interpretation — Work: map statutory and local duties to an operating catalog, onboard teams to appropriate shared services, and define sandbox entry/exit and production gates. Dependencies: program eligibility, available training/evaluation capacity, procurement vehicles, and local implementation resources. Ownership: shared-service leaders maintain common capabilities; local program owners accept use-case risk and outcomes; security/procurement teams own the relevant controls and terms. Skills and adoption: combine awareness training with practical inventory, evaluation, and escalation exercises. Governance checkpoints: intake, risk tiering, sandbox review, procurement, and production transition. Proposed acceptance: the pilot has a current inventory record, trained owners, traceable evaluation evidence, and explicit shared/local operating responsibilities before release. Risks include counting participation as benefit and allowing shared-service availability to obscure local accountability."},"retrievedAt":null,"enrichedAt":"2026-09-05T02:33:27.019Z","enrichmentBasis":"archived evidence"}}]}