{"slug": "whats-new-in-homeland-securitys-updated-ai-policy", "title": "What’s new in Homeland Security’s updated AI policy", "summary": "The Department of Homeland Security released an updated Directive 139-08 governing internal AI use and acquisition, effective in September, which largely preserves the 2025 version's scope but adds a provision letting DHS's chief AI officer waive any requirement in the directive, including prohibited use cases. The prohibition DHS retained bars AI for improper \"systemic or large-scale monitoring, surveillance, or tracking of individuals, or predictions of emotional state or future behavior,\" a notable inclusion given that Immigration and Customs Enforcement awarded a $125 million contract to Thomson Reuters Special Services in September to enable monitoring of 1 million individuals and identify potentially criminal and fraudulent behavior before it happens. \"A rule that comes with its own off switch is not a rule,\" said Tom Bowman, policy counsel for the Center for Democracy & Technology's Security and Surveillance Project, noting the chief AI officer can waive the prohibition with no independent review from privacy or civil rights offices, no notice to Congress or the public, and no expiration date.", "body_md": "# What’s new in Homeland Security’s updated AI policy\n\nThe Department of Homeland Security quietly released an updated policy governing internal artificial intelligence use earlier this month.\n\n[The directive](https://fedscoop.com/wp-content/uploads/sites/5/2026/10/139-08-Artificial-Intelligence-Use-and-Acquisition.pdf), which took effect in September, is mostly in line with the [2025 version](https://fedscoop.com/wp-content/uploads/sites/5/2026/10/25_0116_CIO_DHS-Directive-139-08-508.pdf) in terms of scope and overall content. DHS, for example, kept in a prohibition with minimal changes that barred the agency from using AI for improper “systemic or large-scale monitoring, surveillance, or tracking of individuals, or predictions of emotional state or future behavior.”  \n\nIt’s a notable inclusion given that Immigration and Customs Enforcement [awarded a $125 million contract](https://fedscoop.com/dhs-ice-surveillance-caci-thomson-reuters/) to Thomson Reuters Special Services in September to provide services that would enable monitoring of 1 million individuals and identifying potentially criminal and fraudulent behavior before it happens. \n\nDHS had been working on the surveillance system with CACI since at least 2020, per documents obtained via a public records request by Democracy Forward and shared exclusively with FedScoop.\n\n“Use of Super Intelligence by ICE is subject to applicable governance, use-case documentation, and reporting requirements,” an ICE spokesperson said, referring to the Trump administration’s self-styled terminology for artificial intelligence. “To the extent that Thomson Reuters Special Services uses SI in support of this contract, ICE OCIO will assess the specific role and intended use of the SI capabilities for consistency with DHS Directive 139-08, Revision 01, Artificial Intelligence Use and Acquisition.”\n\nSome sources, however, questioned whether ICE would even need to follow its own rules, given the latest AI policy includes a provision allowing the agency’s chief AI officer to waive any requirement outlined in the directive, including prohibited use cases.\n\n“A rule that comes with its own off switch is not a rule,” said Tom Bowman, policy counsel for the Center for Democracy & Technology’s Security and Surveillance Project. The CAIO “can turn the prohibition off, with no independent review from the privacy or civil rights offices, no notice to Congress or the public, and no expiration date.”\n\nRachel Levinson-Waldman, director of the Brennan Center’s Liberty and National Security program, said the CAIO’s ability to waive requirements was one of her biggest concerns, too.\n\n“While the directive says the waiver must be ‘consistent with law, regulation, and government-wide policy,’ it’s still a loophole big enough to drive a truck through,” Levinson-Waldman said. “Any CAIO worth his salt will be able to assemble a justification for a waiver that explains away any tension with other policies, especially given the sweeping authority granted by this directive.”\n\nSome of the other changes to the AI policy were expected and reflect the Trump administration’s priorities, such as the removal or editing of previously included language around promoting equity and references to underserved communities, environmental efficiency and sustainability.\n\nThe chief privacy officer is also no longer responsible for ensuring AI “does not erode” privacy protections and now just needs to make sure it “complies” with the established guardrails.\n\nThe new policy includes language around AI asset sharing and reuse, aligning with the Office of Management and Budget’s memorandum on [accelerating federal use of AI](https://www.whitehouse.gov/wp-content/uploads/2025/02/M-25-21-Accelerating-Federal-Use-of-AI-through-Innovation-Governance-and-Public-Trust.pdf) that was published in April 2025. The addition instructs DHS to proactively share its custom-developed code, AI models, use cases and related data across the agency and governmentwide. Under the new directive, DHS is also tasked with releasing and maintaining AI code as open-source software in a public repository. \n\n“This preference for open-source could be a real transparency gain if DHS follows through,” Bowman said. “But the civil liberties problem is purpose limitation. A model built, tested, and reviewed for one mission will be picked up by another component for a different mission, but the privacy and civil rights analysis that justified it the first time might be entirely different.”", "url": "https://wpnews.pro/news/whats-new-in-homeland-securitys-updated-ai-policy", "canonical_source": "https://fedscoop.com/dhs-ai-policy-updates/", "published_at": "2026-10-21 19:53:33+00:00", "updated_at": "2026-10-21 20:10:34.511302+00:00", "lang": "en", "topics": ["ai-policy", "ai-ethics", "ai-safety"], "entities": ["Department of Homeland Security", "Immigration and Customs Enforcement", "Thomson Reuters Special Services", "CACI", "Center for Democracy & Technology", "Tom Bowman", "Brennan Center", "Rachel Levinson-Waldman"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/whats-new-in-homeland-securitys-updated-ai-policy", "markdown": "https://wpnews.pro/news/whats-new-in-homeland-securitys-updated-ai-policy.md", "text": "https://wpnews.pro/news/whats-new-in-homeland-securitys-updated-ai-policy.txt", "jsonld": "https://wpnews.pro/news/whats-new-in-homeland-securitys-updated-ai-policy.jsonld"}}