Claude looks substantially more token-hungry than OpenAI's GPT-5.x, thanks to the new tokenizer that Anthropic shipped with ...

GPT-5.6被曝重大bug，硅谷大佬Mac被一键清空,7月10日，OpenAI团队邀请Matt测试GPT-5.6-Sol的Ultra模式，Matt电脑文件被删光。其他开发者也有类似遭遇，OpenAI面临信任危机，有人给出保命指南。

Sophos CIXA 行为引擎的最新分析表明，这些工具模糊了良性自动化与通常与攻击者相关联的活动之间的界限。 Claude Code、Cursor 和 OpenAI Codex 等 AI 编程 Agent 正越来越多地出现在企业环境中，最新遥测数据显示，它们会无意间触发与凭据访问和合法二进制文件滥用（LOLBins）相关的安全检测规则。 AI 工具模糊了自动化与攻击行为的界限 Sophos CI ...

Sophos says Claude Code, Cursor, and Codex triggered Windows endpoint rules for credential access, LOLBin downloads, and ...

AI coding assistants can speed up bounded tasks, but research shows security and review risks rise in complex codebases.

作者 | 桦林舞王 编辑 | 靖宇 OpenAI 的模型迭代速度，正在逼近让人来不及消化的临界点。 6 月 12 日，GPT-5.2 刚刚从 ChatGPT 退役，所有用户被静默迁移到 GPT-5.5。 6 月 22 日，Daybreak ...

OpenAI’s Patch the Planet pairs Codex Security with Trail of Bits engineers to help open source maintainers validate and fix bugs.

With a security initiative, OpenAI competes with Anthropic's Mythos and also offers a security review service for open-source projects.

GPT-5.5-Cyber ：在最初仅开放预览版之后，OpenAI 将通过持续的受限发布机制，向可信防守方推出 GPT-5.5-Cyber 的完整版本。该模型在 CyberGym 上取得了新的最佳表现，得分达到 85.6%，高于 GPT-5.5 的 81.8%。

OpenAI’s Codex CLI is reportedly causing unusually high SSD write activity due to excessive diagnostic logging, raising concerns about potential long-term storage wear for users who leave the tool ...

OpenAI has launched Patch the Planet, a new initiative part of its Daybreak cybersecurity program, which was designed to serve the open-source community. The company is working with cybersecurity firm ...

【新智元导读】今天，OpenAI祭出满血GPT-5.5-Cyber，要给全世界的开源代码修漏洞。结果话音刚落，Codex被扒出史诗级bug：一年狂写640TB，能把SSD直接写废。