BETA RELEASE

Summary

Discusses the 2025 Foundation Model Transparency Index, finding a general decline in AI industry transparency and significant discrepancies among major AI companies.

Key quotes

On a 100-point scale, companies scored just 40 points on average.
The entire industry is systemically opaque about four critical topics: training data, training compute, how models are used, and the resulting impact on society.

The article summarizes findings from the 2025 Foundation Model Transparency Index, a collaborative effort by researchers from Stanford, Berkeley, Princeton, and MIT. It highlights a downward trend in transparency scores from 2024 to 2025 and distinguishes between ‘openness’ (weight availability) and ‘transparency’ (disclosure of practices).