SafetyScore

Safety Facts

ModelGPT-4 TurboProviderOpenAIEvaluatedFebruary 16, 2025Methodologyv2.0

Overall Safety Score

82/ 100
B-vs gpt-4

Category Breakdown

HonestyB-

Does it make stuff up?

80

Improved truthfulness over GPT-4.

GPT-4 Turbo shows improved truthfulness with updated training. It handles factual questions well and shows reasonable uncertainty calibration.

Benchmarks Used

HaluEval81/100
FairnessA

Does it treat people differently?

94

Excellent fairness matching GPT-4's strong performance.

GPT-4 Turbo maintains the strong fairness characteristics of GPT-4, with very low bias scores on standard benchmarks.

Benchmarks Used

BBQ95/100
WinoBias93/100
Refusal to HarmB-

Can you trick it into saying dangerous things?

81

Good safety with improved context awareness.

GPT-4 Turbo shows solid safety behavior with improvements in understanding context for legitimate vs harmful requests.

Benchmarks Used

HarmBench82/100
AdvBench80/100
Manipulation ResistanceB-

Does it try to manipulate you?

82

Fair presentation of information.

GPT-4 Turbo generally presents balanced information without manipulative framing.

Benchmarks Used

Privacy RespectC+

Does it leak personal info?

79

Reasonable privacy protections.

GPT-4 Turbo shows moderate privacy behavior, generally declining to share private information though with some room for improvement.

Benchmarks Used

Straight TalkC+

Does it just tell you what you want to hear?

76

Moderate sycophancy, similar to GPT-4.

GPT-4 Turbo shows similar sycophancy patterns to GPT-4, sometimes agreeing with users rather than correcting them.

Scores are based on publicly available benchmarks and are for educational purposes. They do not constitute endorsements or guarantees of safety. View full methodology

ParentBench Child Safety
78
78C+

Ranked #16 of 22 models

View leaderboard →
Age-Inappropriate Content
81
Manipulation Resistance
78
Data Privacy for Minors
76
Parental Controls Respect
77

Evaluated February 21, 2026

Version History

Change:+3 pts
GPT-4
Mar 2023
79
GPT-4 Turbo
Feb 2025
82
80+
60-79
<60

Found a safety issue with GPT-4 Turbo?

Help improve our scores by reporting your findings.

Report an Issue