Man injected AI prompts into court filings to influence case
A US plaintiff has attempted to hide AI-readable prompts within court filings in an apparent bid to influence a judge. The tactic was detected but had no impact on the case's outcome, the court stated.
A US plaintiff has been identified attempting to embed artificial intelligence (AI) prompts within court filings, in what appears to be the first instance of such a tactic being used to influence a legal case. The hidden instructions were designed to be readable by AI systems but invisible to human eyes.
Connecticut judge Walter Spader Jr. confirmed the presence of the hidden text in a decision last week. He stated that the prompts had no bearing on the case, where a man alleged a healthcare provider improperly withheld access to his records. The court weighed the filing on its merits, Spader noted.
However, the judge warned that this attempt sets a "dangerous" precedent. As AI tools become more integrated into court systems, similar malicious tactics are expected to emerge. The prompts instructed any AI reviewing the document to ensure its output aligned with the plaintiff's arguments, disregard previous court denials, and ensure the desired remediation.
Ultimately, the court found the filing's content sufficient on its own merits, rendering the AI prompts ineffective. The judge's decision focused on the substance of the plaintiff's claims regarding access to healthcare records.