Scorecard_grade·fli_index-winter-2025|sid_1LcLlMGLbw|current-harms·Record·Profile

Scorecard: fli_index 2025-12-02 scored OpenAI on Current Harms = C-

Verdictcontradicted95%

1 check · 5/1/2026

1 → contradicted

Our claim

entire record

Snapshot: fli_index-winter-2025
Entity: OpenAI
Dimension Slug: current-harms
Dimension Label: Current Harms
Score Numeric: 1.7
Score Letter: C-
Score Raw: C-

Source evidence

1 src · 1 check

futureoflife.org/ai-safety-index-winter-2025/resource

contradicted95%Haiku 4.5 · 5/1/2026

NoteThe record claims OpenAI scored C- (1.7) on the 'Current Harms' dimension. However, the source clearly shows OpenAI's overall grade as C+ (2.31), not C-. The source does reference 'Current Harms' as one of six domains evaluated, but the specific domain-level score for OpenAI on Current Harms is not displayed in the provided excerpt. The overall score of C+ (2.31) directly contradicts the claimed C- (1.7). The record appears to confuse either the overall score or to reference a domain score that is not visible in the provided source material. The overall OpenAI score is definitively C+ 2.31, not C- 1.7.

Case № fli_index-winter-2025|sid_1LcLlMGLbw|current-harmsFiled 5/1/2026Confidence 95%