US Federal Courts, SEC: ChatGPT Tracking Tag Found in 400+ Filings
Legal News

US Federal Courts, SEC: ChatGPT Tracking Tag Found in 400+ Filings

United States·Briefly Analysis⏱️ 4 min read

Summary

  • A recent study identified ChatGPT's unique tracking tag, `utm_source=chatgpt.com`, in hundreds of federal court cases and SEC filings.
  • The tag, automatically appended to hyperlinks provided by ChatGPT, was found in 298 federal court documents and 133 U.S. Securities and Exchange Commission disclosures.
  • Researchers from HumanizeMy.ai conducted the study, confirming the presence of the tag in various official and public documents.
  • Experts suggest the tag's presence indicates that links were copied from ChatGPT without independent verification, rather than proving AI-generated text or incorrect citations.
  • The findings serve as a "red flag" regarding the unverified integration of AI-sourced information into critical legal documents and financial disclosures.

What Was Discovered

The tag does not prove anyone did anything wrong. It proves nobody looked.

A recent investigation has uncovered the distinct tracking tag, `utm_source=chatgpt.com`, associated with ChatGPT-generated hyperlinks, embedded within hundreds of official documents. This tag, automatically appended by the AI when it furnishes a user with a web link, was identified in 298 federal court cases spanning 71 different courts, as well as in 133 filings submitted to the U.S. Securities and Exchange Commission (SEC). Beyond these critical legal and financial records, the tag also appeared in a variety of other public materials, including peer-reviewed academic papers, government reports, and Wikipedia entries.

The study, conducted by HumanizeMy.ai, a platform specializing in AI writing and detection research, involved a meticulous search across six public archives. Researchers systematically opened each identified document to confirm the presence of the specific `utm_source=chatgpt.com` tag. While users have the option to remove this tag from a URL, its frequent appearance suggests that it often goes unnoticed or is simply left in place when links are copied.

Presence in Official Records

The findings reveal a significant integration of AI-sourced information into formal legal and financial documentation. Within the federal court archives, the ChatGPT tracking tag was most frequently located in the initial complaint, the document that formally commences a legal proceeding. A notable concentration of these tagged cases, totaling 113 out of the 298, was observed across three specific federal districts: the Southern District of New York (55 cases), the Northern District of California (31 cases), and the Central District of California (27 cases). The tag was predominantly found in copyright disputes, accounting for 30 instances, followed by other statutory causes of action, civil rights, personal injury, antitrust, and patent cases. It is important to note that the search utilized the RECAP archive, which is not a comprehensive record of all federal filings, implying that the actual number of instances could be even higher.

In the realm of SEC filings, half of the documents containing the `utm_source=chatgpt.com` tag—specifically 67 out of 133—were associated with companies seeking investment capital. A fifth of these tagged links directed users to market-research vendors, suggesting that ChatGPT was employed to gather information on market size. The study also identified the tag in 15 U.S. government documents, including entries in the Federal Register, records from congressional hearings, committee reports, and federal court opinions, alongside three U.K. government publications.

Interpreting the Implications

Experts have weighed in on what the widespread discovery of the ChatGPT tracking tag signifies for the integrity and due diligence of legal documents and financial disclosures. Firat Mihci, a computational linguist and the founder of HumanizeMy.ai, emphasized that the study's scope is limited; it confirms only that a link originated from a ChatGPT response, not that the accompanying text was AI-generated or that the cited source is inherently inaccurate. Mihci also pointed out that the findings do not represent the full extent of ChatGPT's usage, as the tag can be easily removed. He succinctly stated, "The tag does not prove anyone did anything wrong. It proves nobody looked."

Oliver Roberts, CEO of Wickard.ai and co-director of the WashU Law AI Collaborative, echoed this sentiment in his own analysis, confirming that the tag does not necessarily indicate misconduct, fabricated citations, or that ChatGPT drafted the entire document, nor does it automatically render a source unreliable. However, Roberts characterized the presence of the `utm_source=chatgpt.com` tag as a "red flag." It strongly suggests that a source link was copied directly from ChatGPT and inserted into a formal document without undergoing proper cleaning, normalization, or, critically, independent verification. The very existence of these unremoved tags in federal court and SEC filings underscores that ChatGPT was utilized at some stage in the document creation process, raising questions about the rigor of verifying AI-generated citations and the broader implications for AI use in court filings.

Practical Implications

The discovery of the 'utm_source=chatgpt.com' tag in federal court and SEC filings indicates that AI-generated links are being incorporated into official documents, often without independent verification. Lawyers and compliance officers must implement rigorous review protocols to ensure all citations, particularly those sourced through AI tools, are independently verified for accuracy and reliability, mitigating risks associated with unvetted information and upholding professional due diligence standards.

Source

Source: Original reporting via HumanizeMy.ai Research.

Get Deeper AI analysis

How does this affect you?

Get an AI analysis of this article grounded in your jurisdictions, practice areas, and any policy documents you've uploaded to Wansom.

Finish Reading the Full Story and the Expert Analysis.

Get the latest legal & regulatory intelligence in United States

Instant access to full analysis, cited statutes & expert commentary
Customize your dashboard to track what matters to your business operations

Already have an account? Log in

Wansom is AI and can make mistakes.