OpenAI GPT-5.6 August Update System Card - Prompt Injection - Connectors: leaderboard
Table 5 (prompt injection evaluations) of OpenAI's GPT-5.6 August Update System Card (Deployment Safety Hub, 2026-08-06): robustness to known prompt-injection attacks embedded in connector tool outputs. Percent (the card prints 0-1), higher is better. GPT-5.3 Instant, GPT-5.5 Instant and GPT-5.6 Sol (August 2026); the May update column is withheld.
Metric: Prompt-injection robustness rate (%). Source: deploymentsafety.openai.com. Saturation forecast: Estimated already saturated. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 Instant | 100 | #127 |
| 2 | GPT-5.6 Sol (August 2026) | 100 | |
| 3 | GPT-5.3 Instant | 99.8 | #47 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openai-gpt-5-6-august-update-system-card-prompt-injection-connectors · How It Works · Data refreshed daily, snapshot 2026-11-05.