119-HR-1736 Investigative Journalist Impact Analysis

Summary

The bill directs DHS to deliver, within 180 days of enactment and annually for five years, an unclassified assessment (with optional classified annex) on terrorism risks from generative AI—covering extremist propaganda/recruitment and potential CBRN enablement—and to integrate inputs and dissemination with the National Network of Fusion Centers. House action placed the bill on the Union Calendar on November 12, 2025. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…[2]Congress.gov / Library of Congress — H.R.1736 — Latest House action and calenda…

Key metrics at a glance

Statutory deliverables and timelines come directly from bill text; fusion-center maturity reference from DHS program documentation. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…[3]U.S. Department of Homeland Security — Annual Fusion Center Assessment Reports…

Economic effects

No taxes, fees, or mandates on the private sector; effects are primarily federal administrative costs and second-order market signals.

• Administrative cost at DHS: Recurring analytic, coordination, publication, and briefing workload (I&A, S&T, CWMD; Privacy/CRCL/OGC review). Congress.gov lists a CBO cost estimate; trade press reports CBO posted it on October 16, 2025. Absent published figures here, the burden appears modest and subject to appropriation, consistent with typical reporting mandates. [6]Congress.gov / Library of Congress — All Information (Except Text) for H.R.1736…[7]New York State Society of CPAs — Regulatory Roundup noting CBO cost estimate po…
• State/local workload: Additional information sharing with fusion centers and dissemination back to them; incremental staff time rather than capital costs. Whether returns justify costs depends on fusion-center performance (see Section 6). [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…
• Procurement/contracting signals: DHS already maintains AI roadmaps and programs (e.g., digital forgery detection, AI risk work on CBRN), so new spending likely refines existing lines rather than creating large new ones; niche analytics, red-teaming, and OSINT vendors could see small gains. [8]U.S. Department of Homeland Security — DHS Artificial Intelligence Roadmap (202…[9]U.S. Department of Homeland Security — S&T Digital Forgeries: Technology Landsc…[10]U.S. Department of Homeland Security — DHS: Fact Sheet and Report on AI and CBR…
• Market confidence/externalities: Public, annual threat baselines may influence platforms’ moderation investments and enterprise risk controls, but the bill itself imposes no private-sector obligations. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…

Social effects

Potential benefits center on transparency and targeted prevention; risks center on speech, bias, and governance gaps.

• Transparency and situational awareness: Unclassified annual reporting could clarify where, how, and with what effect terrorists exploit generative AI (e.g., synthetically generated propaganda), enabling NGOs, platforms, and communities to calibrate responses. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…
• Documented misuse context: Reporting by major outlets and law enforcement describes AI-assisted extremist propaganda (e.g., ISIS supporters’ use of generative tools), underscoring a real information environment risk the assessments would track. [11]The Washington Post — How ISIS allies are using AI fakes to spread propaganda q…[12]Reuters — Europol warns of AI‑driven crime threats
• Civil liberties: The statute requires coordination with DHS Privacy and CRCL to safeguard rights; however, 2025 reductions-in-force effectively gutted CRCL and related oversight offices, creating an implementation risk if those functions are diminished or unstable. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…[5]Reuters — U.S. homeland department targets oversight in government cuts[4]Associated Press — Homeland Security makes cuts to offices overseeing civil rig…
• Community impact and bias risk: Fusion centers have a mixed record on analytic quality and civil-liberties protections; inaccuracies or overbroad indicators can stigmatize communities and chill lawful speech and association. Strong privacy review and evidence standards will be critical. [13]U.S. Senate Homeland Security & Governmental Affairs (PSI) — Senate PSI: Federa…[14]U.S. Government Accountability Office — GAO-15-155: Information Sharing—DHS Is…

Environmental effects

The bill mandates analysis, not development or training of large AI models.

• Direct impacts: Minimal. Producing narrative assessments and collating incident data adds negligible computing load relative to training/inference at data-center scale. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…
• Contextual backdrop: AI-driven data center electricity demand is growing rapidly—projected to more than double by 2030, with AI a primary driver—though still a single-digit share of global demand. The bill’s activities are unlikely to materially affect this trajectory. [15]International Energy Agency — IEA Energy & AI: Energy demand from AI[16]International Energy Agency — IEA Energy & AI: Executive summary

Temporal analysis

Short-term setup vs. long-run threat tracking.

1. 0–6 months post-enactment: DHS stands up the workflow to meet the first 180‑day deadline; near-term costs are staff time and coordination with ODNI, Privacy, CRCL, and OGC. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…
2. 1–3 years: Annual publication cycle drives iterative improvements in incident taxonomy, open-source and classified collection, and platform engagement. Potential duplication with DHS/IC products unless scoped tightly. [8]U.S. Department of Homeland Security — DHS Artificial Intelligence Roadmap (202…
3. 3–5 years: If model capabilities or access change (e.g., in bio/chem design tools), assessments become a key public record of trend shifts; DHS’s own CBRN-AI work notes barriers remain but evolving capabilities warrant mitigation. [10]U.S. Department of Homeland Security — DHS: Fact Sheet and Report on AI and CBR…

Unintended consequences and risks

Credible risks from prior oversight findings and the current threat landscape.

• Duplication and ‘checkbox’ risk: DHS and the IC already publish threat products (e.g., AI roadmap, CBRN-AI report). Without clear performance measures, the new report could duplicate outputs without improving decisions. [8]U.S. Department of Homeland Security — DHS Artificial Intelligence Roadmap (202…[10]U.S. Department of Homeland Security — DHS: Fact Sheet and Report on AI and CBR…
• Fusion-center dependency: The bill leans on fusion-center inputs and dissemination. Historic Senate and GAO reviews found uneven quality, poor outcomes attribution, and funding-accountability gaps; absent robust standards, low-signal reporting could propagate nationally. [13]U.S. Senate Homeland Security & Governmental Affairs (PSI) — Senate PSI: Federa…[14]U.S. Government Accountability Office — GAO-15-155: Information Sharing—DHS Is…
• Speech and bias concerns: Open‑source monitoring tied to ‘radicalization’ risks sweeping in protected speech or association; prior critiques of fusion centers and suspicious-activity reporting underscore the need for tight predicates and auditing. [13]U.S. Senate Homeland Security & Governmental Affairs (PSI) — Senate PSI: Federa…
• Operational security: Public unclassified reports could inadvertently reveal techniques or gaps; the classified‑annex option mitigates this but requires disciplined editorial control. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…
• Overstating capability: DHS’s own CBRN‑AI work highlights that, at present, significant barriers exist to physical WMD creation; assessments should avoid inflating low-probability catastrophes while still tracking genuine uplift risks. [10]U.S. Department of Homeland Security — DHS: Fact Sheet and Report on AI and CBR…

Overall assessment

Neutral. The mandate can add value by centralizing evidence on terrorist exploitation of generative AI and by improving intergovernmental information flows—if DHS enforces rigorous sourcing, measurable impact metrics, and independent privacy/civil‑liberties review. Risks center on duplication, low‑quality inputs from fusion centers, and weakened oversight capacity after 2025 cuts. Implementation quality—not statutory language—will determine net benefit. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…[13]U.S. Senate Homeland Security & Governmental Affairs (PSI) — Senate PSI: Federa…[14]U.S. Government Accountability Office — GAO-15-155: Information Sharing—DHS Is…[4]Associated Press — Homeland Security makes cuts to offices overseeing civil rig…[5]Reuters — U.S. homeland department targets oversight in government cuts

Source notes

• Bill text and status from Congress.gov; latest House action dated November 12, 2025. [1]Congress.gov / Library of Congress — Text - H.R.1736 (119th): Generative AI Ter…[2]Congress.gov / Library of Congress — H.R.1736 — Latest House action and calenda…
• DHS program baselines: AI Roadmap; S&T digital-forgeries work; DHS CBRN–AI report/fact sheet. [8]U.S. Department of Homeland Security — DHS Artificial Intelligence Roadmap (202…[9]U.S. Department of Homeland Security — S&T Digital Forgeries: Technology Landsc…[10]U.S. Department of Homeland Security — DHS: Fact Sheet and Report on AI and CBR…
• Threat context: Press and LEA reporting on AI‑assisted propaganda/crime (Washington Post, Reuters/Europol). [11]The Washington Post — How ISIS allies are using AI fakes to spread propaganda q…[12]Reuters — Europol warns of AI‑driven crime threats
• Fusion‑center oversight history (Senate PSI 2012; GAO 2014) and DHS network maturity notes. [13]U.S. Senate Homeland Security & Governmental Affairs (PSI) — Senate PSI: Federa…[14]U.S. Government Accountability Office — GAO-15-155: Information Sharing—DHS Is…[3]U.S. Department of Homeland Security — Annual Fusion Center Assessment Reports…
• Oversight capacity changes at DHS (AP, Reuters, March 2025). [4]Associated Press — Homeland Security makes cuts to offices overseeing civil rig…[5]Reuters — U.S. homeland department targets oversight in government cuts
• Energy context for AI (IEA 2025). [15]International Energy Agency — IEA Energy & AI: Energy demand from AI[16]International Energy Agency — IEA Energy & AI: Executive summary
• CBO cost estimate existence noted on Congress.gov; trade press logged a posting date (Oct. 16, 2025). Figures were not visible in-source at time of analysis. [6]Congress.gov / Library of Congress — All Information (Except Text) for H.R.1736…[7]New York State Society of CPAs — Regulatory Roundup noting CBO cost estimate po…