What happened
Partnership on AI evaluated 13 organizations on public documentation of foundation model impacts, reviewing over 150 papers, articles, and reports. The assessment measures progress across four core practices: pre-deployment disclosure, post-deployment monitoring, impact assessment, and stakeholder engagement. The report builds on PAI's 2023 Guidance for Safe Foundation Model Deployment and 2025 progress-tracking methodology. Findings reveal that as foundation model providers increasingly share information privately, public transparency on impacts has declined, with organizations retreating from earlier commitments to open disclosure even as models shape critical societal systems.
Why it matters
This report quantifies a troubling trend: model providers are sharing less publicly at precisely the moment when their systems are being integrated into healthcare, finance, education, and government operations. For boards and CISOs, the transparency gap complicates risk assessment, vendor evaluation, and compliance planning when deploying AI systems.
Action needed
Review your vendor AI disclosure requirements against PAI's transparency framework. Escalate to procurement and risk committees if current AI vendors score poorly on public impact documentation. Consider requiring third-party transparency assessments as a condition of AI procurement contracts.