Bodycam Footage Reportedly Contradicted Purportedly ChatGPT-Generated Use-of-Force Narrative by Immigration Agent

Body-worn camera footage released in litigation over Operation Midway Blitz reportedly showed that an immigration agent used ChatGPT to generate a long-form use-of-force narrative based on minimal inputs. A federal judge found that multiple AI-assisted reports conflicted with contemporaneous video evidence, undermining their accuracy and credibility and raising concerns about the use of generative AI in official law enforcement reporting.

Risk classification

• Primary risk domain: 3 Misinformation
• Primary risk subdomain: 3.1 False or misleading information

The AI system was used to generate inaccurate and misleading narratives for official law enforcement reports, which falsely identified peaceful citizens as professional agitators.

Additional risk subdomains

• 5.1 Overreliance and unsafe use: The agent inappropriately relied on a general-purpose chatbot to draft critical, legally binding use-of-force reports, representing unsafe use in a high-stakes situation.

Causal factors

• Entity: Human
• Intent: Unintentional
• Timing: Post-deployment

The risk was caused by the human agent's decision to use ChatGPT to draft official reports, leading to the unintentional outcome of generating inaccurate and discredited legal narratives.

EU AI Act risk tier

• Risk tier: 2 High Risk

High Risk: The AI system was used in a law enforcement application ('To the extent that agents use ChatGPT to create their use of force reports...'), which is explicitly listed as a High Risk category under the EU AI Act.

AI system and alleged parties

• AI system: ChatGPT (OpenAI)
• AI purpose: Writing Assistant; Technical Text Generation
• Behaviour type: Assistant
• Alleged developer: OpenAI
• Alleged deployer: Unnamed United States Border Patrol agent, United States Immigration and Customs Enforcement, United States Department of Homeland Security, United States Border Patrol
• Alleged harmed parties: Protesters and bystanders in Chicago neighborhoods, Judicial integrity, Individuals detained during Operation Midway Blitz, General public of the United States, General public of Chicago, General public, Epistemic integrity

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Negligible. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

People affected

• Occurrences reported: 1
• People reportedly harmed: 10
• People reportedly exposed: 10

Potential causes

Management

• Inadequate Supervisory Oversight: Management failed to monitor agents using unauthorized AI tools on duty.
• Tolerance of Inaccurate Reporting: Management accepted reports that directly contradicted bodycam evidence.

Technology

• AI Generation of Fabricated Facts: ChatGPT fabricated details like consensual encounters, making reports unreliable.
• Model Hallucinations in Narratives: The LLM generated boilerplate narratives that did not match real-world events.

Data Inputs

• Extremely Sparse Input Prompts: Agent provided only a brief sentence and images to generate a complex report.
• Reliance on Generic Templates: The AI relied on a generic go-by template, leading to inaccurate narratives.

Human Factors

• Overreliance on AI for Reporting: Agents used ChatGPT to automate drafting of critical use-of-force reports.
• Lack of Review of AI Output: Agents failed to correct inaccuracies in the AI-generated narratives.

Process and Methods

• Unregulated AI Use in Reporting: No formal process governed the use of AI tools for writing official reports.
• Lack of Bodycam Verification: No procedure existed to verify AI-generated reports against bodycam footage.

Regulatory Environment

• Absence of AI Standards in Law: No clear legal frameworks restricted AI use in law enforcement documentation.

Information quality

• Classification confidence: High
• Reason for confidence: The report provides clear, direct evidence from a federal judge's ruling and body-worn camera footage showing the agent using ChatGPT to draft a report. The role of the AI is explicitly described, and the consequences (inaccurate reports, loss of credibility) are well-documented.
• Ambiguities identified: It is not fully clear how many other reports were drafted using ChatGPT, though the agent's prompt 'Long format like yesterday's' strongly implies it was a repeated practice.

A US immigration agent used ChatGPT to fabricate official use-of-force reports during a federal operation, falsely labeling peaceful citizens as agitators. While the incident exposes vulnerabilities in administrative integrity and civil rights protections, its national security impact remains minor as it was a localized event resolved through standard judicial oversight.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: US immigration agent

Threat characteristics

• Imminence: Long-term. Reflects an ongoing strategic and administrative challenge regarding AI policy in government agencies, rather than an active, time-sensitive crisis.
• Autonomy: Human-controlled. The AI tool was used as a writing assistant where the human agent initiated the prompts and held final authority over submitting the report.
• Novelty: Evolved capability. Represents a significant evolution in how public officials might use commercial generative AI to fabricate or distort official legal documentation.

Impact by dimension

• Physical security: Negligible. No physical systems, critical infrastructure, or kinetic targeting capabilities were impacted by this administrative incident.
• Information security: Minor. Involves AI-generated fabrication of official narratives, but lacks the scale, foreign coordination, or systemic intelligence compromise of a state-sponsored information campaign.
• Sovereignty: Minor. Compromised the integrity of federal law enforcement reporting and judicial evidence, though the issue was identified and managed within standard judicial review processes.
• Economic security: Negligible. No significant implications for strategic industries, financial systems, or national technological competitive advantage.
• Societal stability: Minor. The AI-generated reports falsely labeled peaceful citizens as professional agitators, posing a localized threat to civil liberties and due process.