GPT-6 Astra hallucinates less but remains vulnerable to hidden prompt injection attacks
Original title:OpenAI's GPT-6 Astra hallucinates less but remains vulnerable to hidden prompt injections
AI summary
The Decoder reported that OpenAI's new GPT-6 Astra hallucinates less than predecessor GPT-5.6 Sol and defends against direct prompt injection 99.99% of the time, but its defense rate drops to about 67% under multi-turn adaptive attacks.
Key Takeaways
- The Decoder reported that OpenAI's new GPT-6 Astra hallucinates less than predecessor GPT-5.
- 6 Sol and defends against direct prompt injection 99.
- 99% of the time,but its defense rate drops to about 67% under multi-turn adaptive attacks.
Quick Facts
- Published
- Updated
- Source
- The Decoder:AI News(RSS)
- Category
- Tutorials & Tips
- AI score
- 84
文章汇总 OpenAI 系统卡与 Gray Swan 独立测试数据,读者可据此比较 GPT-6 Astra 与竞品在幻觉和注入攻击上的实际防线。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: GPT-6 Astra hallucinates less but remains vulnerable to hidden prompt injection attacks. September 5, 2026. https://new.nat6.net/en/news/cmtn8fc1w0qb4romyobllbzv9/