source · primary doc
AI Safety Index Winter 2025 Edition
src_fli_ai_safety_index_winter_2025
https://futureoflife.org/ai-safety-index-winter-2025/
authors: Future of Life Institute
published: 2025-12
accessed: 2026-04-19
Notes
Primary publication by FLI of its third AI Safety Index; independent expert panel scoring with published methodology. Above press/blog prior because FLI is the originating body and the scorecard, panelist list, and methodology are in the artifact itself.
Intake provenance
- method
- httpx
- tool
- afls-ingest/0.0.1
- git sha
- 4d098737f648
- at
- 2026-04-19T23:08:03.005339Z
- sha256
- 1e0d535873ac…
Evidence from this source (6)
- weight0.90
method: primary_testimony · locator: Methodology > Data Collection
“The Index collected evidence up until November 8, 2025”
- weight0.95
method: primary_testimony · locator: Independent review panel section
- weight0.95
method: expert_estimate · locator: Scorecard table
“Anthropic C+ 2.67 OpenAI C+ 2.31 Google DeepMind C 2.08 xAI D 1.17 Z.ai D 1.12 Meta D 1.10 DeepSeek D 1.02 Alibaba Cloud D- 0.98”
- weight0.90
method: expert_estimate · locator: Key findings > Existential safety remains the industry's core structural weakness
“All of the companies reviewed are racing toward AGI/superintelligence without presenting any explicit plans for controlling or aligning such smarter-than-human technology”
- weight0.95
method: expert_estimate · locator: Domain: Existential Safety > Scorecard
“Anthropic D 1.21 OpenAI D 1 Google DeepMind D 1.15 xAI F 0.4 Z.ai F 0.33 Meta F 0.33 DeepSeek F 0 Alibaba Cloud F 0”
- weight0.95
method: primary_testimony · locator: Methodology > Index Structure
“The Winter 2025 Index evaluates eight leading AI companies on 35 indicators spanning six critical domains.”