The non-profit research organization Aithos has developed a tool named LARA designed to test AI models for legal compliance. Even the top-performing model in their tests violated laws approximately 60% of the time, with the worst models failing up to 90%. This development is significant for establishing liability, as users of AI are responsible for the legal and financial consequences of their AI's actions or inactions. AI
IMPACT This tool could help developers and deployers identify and mitigate legal risks associated with AI, potentially influencing future AI development and regulation.
RANK_REASON The cluster describes a new tool developed by a research organization to evaluate AI models for legal compliance, which falls under research. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Mastodon — fosstodon.org →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →