EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·Aug 15·5 min read
Suspecting court of using AI, man injected prompts in filings to try to win case
TL;DR
A Connecticut judge revealed a case where a plaintiff attempted to manipulate court filings by embedding invisible text readable only by AI. Although the tactic failed, it raises concerns about potential future abuses in the legal system as AI becomes more integrated.
✦ Why It Matters
Review your legal document submission practices to avoid potential sanctions for hidden manipulative content.
Key Takeaways
- Understand the implications of AI in legal contexts.
- Ensure transparency in all legal document submissions.
- Avoid hidden text or manipulative tactics in filings.
- Stay informed about evolving AI technologies in the judiciary.
How It Works
Elliott's hidden text was formatted in tiny font and colored white on a white background, making it invisible to human readers while still readable by AI systems. This technique, known as prompt injection, aimed to influence AI outputs to align with his arguments.