Author: RadarAI Editorial
Editor: RadarAI Editorial
Last updated: 2026-11-05
Review status: Editorial review pending
Brief
速报
官方
AI动态
开源
OpenAI went all out on both the image and security fronts this week: ChatGPT Images 2.6 launched with GIF animation and reference image features, while on the developer side GPT-Image-2.6 Flare/Sunburst boosted generation speed by 50%, producing images in as fast as six seconds [1][2][4]. Meanwhile, its autonomous vulnerability discovery pipeline reproduced 1,724 vulnerabilities in a single run and generated 948 Linux kernel privilege escalation exploits, at an average cost of about $1,279 per vulnerability, ...
Editorial standards and source policy: Editorial standards, Team. Content links to primary sources; see Methodology.
## 🔍 Key Insights
**OpenAI** went all out on both the image and security fronts this week: **ChatGPT Images 2.6** launched with GIF animation and reference image features, while on the developer side **GPT-Image-2.6 Flare/Sunburst** boosted generation speed by 50%, producing images in as fast as six seconds [1][2][4]. Meanwhile, its autonomous vulnerability discovery pipeline reproduced **1,724 vulnerabilities** in a single run and generated **948 Linux kernel privilege escalation exploits**, at an average cost of about **$1,279** per vulnerability, demonstrating the real-world capabilities of frontier models in the cybersecurity domain [10][11][12][13].
## 🚀 Top Stories
- **OpenAI releases Images 2.6 and opens Codex to all tiers** [1]: Available across all ChatGPT platforms, with Codex covering web, desktop, and mobile.
- **GPT-Image-2.6 Flare and Sunburst arrive on the API** [2]: Flare generates in as fast as six seconds, while Sunburst focuses on high quality and cross-edit consistency, with SVG export support.
- **ChatGPT Images 2.6 supports animated GIF generation** [3]: Prompts or images can be turned into animated GIFs, with speed adjustment, single-frame editing, and loop sharing.
- **OpenAI API supports annotated image editing** [5]: Users can send annotated images to precisely specify edit locations and set image quality separately based on detail requirements.
- **OpenAI agents discover at least 10 new Linux exploit techniques** [10]: To its and experts' knowledge, all were previously unknown; the model can discover and patch new vulnerabilities rather than applying known techniques.
- **Autonomous vulnerability pipeline generates 948 privilege escalation exploits in a single run** [11][13]: Of 2,308 candidate findings, 1,724 were verified, with total API costs of about $1.2 million.
- **Apple tightens employee AI usage budgets** [14]: Daily limits for some engineers cut from $300 to $200, with non-engineering teams potentially squeezed to $25.
- **webAI releases 8.8B reasoning model TwIL-8B** [21]: Post-trained on IBM Granite 4.2 8B, scoring 98.2 on MATH-500, with weights released on Hugging Face under a non-commercial license.
## 🔗 Sources
[1] OpenAI releases Images 2.6 and opens Codex to all tiers — https://aihot.news/items/o915zd1u1eq13gyo9x3nsgoq8
[2] OpenAI releases GPT-Image-2.6 Flare and Sunburst developer image models — https://aihot.news/items/v4926jbeulchfze584bf3em1r
[3] ChatGPT Images 2.6 supports animated GIF generation — https://aihot.news/items/lgg076xbmrpmfpdoi1xa50pf0
[4] OpenAI releases ChatGPT Images 2.6, adding GIF and reference image features — https://aihot.news/items/jkz26nto61yp9panc1o0bj38s
[5] OpenAI API supports annotated image editing — https://aihot.news/items/c2uwzop94yqfdc0qjnzl67zag
[10] OpenAI says its agents discovered at least 10 new Linux exploit techniques — https://aihot.news/items/ulr0gn1mg4kpj7yz2h2zwp7u1
[11] OpenAI demonstrates model reproducing 1,724 vulnerabilities and generating 948
OpenAI went all out on both the image and security fronts this week: ChatGPT Images 2.6 launched with GIF animation and reference image features, while on the developer side GPT-Image-2.6 Flare/Sunburst boosted generation speed by 50%, producing images in as fast as six seconds [1][2][4]. Meanwhile, its autonomous vulnerability discovery pipeline reproduced 1,724 vulnerabilities in a single run and generated 948 Linux kernel privilege escalation exploits, at an average cost of about $1,279 per vulnerability, demonstrating the real-world capabilities of frontier models in the cybersecurity domain [10][11][12][13].
🚀 Top Stories
- OpenAI releases Images 2.6 and opens Codex to all tiers [1]: Available across all ChatGPT platforms, with Codex covering web, desktop, and mobile.
- GPT-Image-2.6 Flare and Sunburst arrive on the API [2]: Flare generates in as fast as six seconds, while Sunburst focuses on high quality and cross-edit consistency, with SVG export support.
- ChatGPT Images 2.6 supports animated GIF generation [3]: Prompts or images can be turned into animated GIFs, with speed adjustment, single-frame editing, and loop sharing.
- OpenAI API supports annotated image editing [5]: Users can send annotated images to precisely specify edit locations and set image quality separately based on detail requirements.
- OpenAI agents discover at least 10 new Linux exploit techniques [10]: To its and experts' knowledge, all were previously unknown; the model can discover and patch new vulnerabilities rather than applying known techniques.
- Autonomous vulnerability pipeline generates 948 privilege escalation exploits in a single run [11][13]: Of 2,308 candidate findings, 1,724 were verified, with total API costs of about $1.2 million.
- Apple tightens employee AI usage budgets [14]: Daily limits for some engineers cut from $300 to $200, with non-engineering teams potentially squeezed to $25.
- webAI releases 8.8B reasoning model TwIL-8B [21]: Post-trained on IBM Granite 4.2 8B, scoring 98.2 on MATH-500, with weights released on Hugging Face under a non-commercial license.
🔗 Sources
[1] OpenAI releases Images 2.6 and opens Codex to all tiers — https://aihot.news/items/o915zd1u1eq13gyo9x3nsgoq8
[2] OpenAI releases GPT-Image-2.6 Flare and Sunburst developer image models — https://aihot.news/items/v4926jbeulchfze584bf3em1r
[3] ChatGPT Images 2.6 supports animated GIF generation — https://aihot.news/items/lgg076xbmrpmfpdoi1xa50pf0
[4] OpenAI releases ChatGPT Images 2.6, adding GIF and reference image features — https://aihot.news/items/jkz26nto61yp9panc1o0bj38s
[5] OpenAI API supports annotated image editing — https://aihot.news/items/c2uwzop94yqfdc0qjnzl67zag
[10] OpenAI says its agents discovered at least 10 new Linux exploit techniques — https://aihot.news/items/ulr0gn1mg4kpj7yz2h2zwp7u1
[11] OpenAI demonstrates model reproducing 1,724 vulnerabilities and generating 948
← Back to Updates