LatticeFlow launches evaluation framework for compliance with EU AI Act
Summary: LatticeFlow, ETH Zurich, and INSAIT have launched Compl-AI, the first evaluation framework for compliance with the EU AI Act. This framework aims to clarify how AI models meet the Act's ethical principles, which have been criticized for being vague.
Compl-AI evaluates models from major companies like OpenAI and Google across 27 technical areas. The framework highlights shortcomings in diversity, non-discrimination, and fairness, with no model achieving perfect compliance. GPT-4 Turbo and Claude 3 Opus scored highest at 0.89.
The EU AI Act categorizes models by risk levels, with serious penalties for non-compliance. Compl-AI's launch is seen as a step toward translating the Act's requirements into measurable standards, with regulators expressing support for the initiative.
This is article metrics. Combined, they form a significance score, that indicates how important the news is on a scale from 0 to 10.
My algorithm scores 10,000 news articles daily, and creates a single significance-ordered list of news.
Read more about how I calculate significance, or see today's top ranked news on the main page:
See today's news rankings