Signal is active with 1 mentions in latest observed window
Recommended action window: Monitor and triage in normal cycle
NVD description
An integer overflow was addressed with improved input validation. This issue is fixed in iOS 18.7.9 and iPadOS 18.7.9, macOS Sequoia 15.7.7, macOS Sonoma 14.8.7, macOS Tahoe 26.5. An app may be able to cause unexpected system termination.
Priority
LOW
Exploitation
NONE
PoC
NONE
Patch
AVAILABLE
Momentum
STABLE
If you run products in this scope, you should treat this CVE as relevant to your environment.
| Date | Total | Labels |
|---|
| 2026-05-26 | 19 | Disclosure13General4Patch2 |
| 2026-05-29 | 1 | Disclosure1 |
Claude找出 macOS 内核洞，已经进 Apple 安全公告了。 macOS Tahoe 26.5 修了 CVE-2026-28952：Kernel 整数溢出，app 可导致系统异常终止，署名 Calif_io、Claude 和 Anthropic Research。 安全行业的分工在变：AI 不再只写 PoC，开始进漏洞署名栏。拐点在这一行。
Post summary
Apple announced the patch for CVE-2026-28952, a kernel integer overflow discovered by AI, but no exploit or active use in the wild is reported.
Apple 官方发布安全公告：CVE-2026-28952 macOS 26.5 内核漏洞——发现者是 Claude。 意思是： Anthropic 的 Claude 第一次以"独立漏洞研究员"身份被 Apple 记入安全公告。 AI 不只是给苹果写测试用例，是在替苹果找内核级别的漏洞： 这是 2026 年 AI 安全研究的分水岭事件。 对应的产业变化： 安全研究员这个职业，正在被 AI 部分替代。 但同时也被 AI 放大——会用 Claude 做安全研究的人，效率是 10 倍以上。 整个安全行业的"门槛"和"天花板"同时被改写。
Post summary
Apple announced CVE-2026-28952, a kernel vulnerability in macOS 26.5 discovered by Anthropic’s Claude. The incident highlights AI's growing role in uncovering security flaws.
Apple 刚发的安全公告，CVE-2026-28952，macOS 26.5 内核漏洞。 是 Claude 发现的。 不是"AI 辅助安全研究员发现"。是 AI 找到的。 顶尖安全研究员花职业生涯磨练的能力，AI 现在开始系统性地做了。
Post summary
Apple has released a security advisory for CVE-2026-28952, a macOS 26.5 kernel vulnerability, noting that it was found by AI.
🔍 𝗔𝗽𝗽𝗹𝗲 𝗖𝗿𝗲𝗱𝗶𝘁𝘀 𝗖𝗹𝗮𝘂𝗱𝗲 𝗳𝗼𝗿 𝗙𝗶𝗻𝗱𝗶𝗻𝗴 𝗧𝘄𝗼 𝗺𝗮𝗰𝗢𝗦 𝗩𝘂𝗹𝗻𝗲𝗿𝗮𝗯𝗶𝗹𝗶𝘁𝗶𝗲𝘀 Apple's macOS Tahoe 26.5 security update credits Claude and Anthropic for discovering two CVEs. CVE-2026-28952: an integer overflow found by http://Calif.io with Claude and Anthropic Research. CVE-2026-28942: a WebKit vulnerability found by Milad Nasr and Nicholas Carlini with Claude. Real-world proof that AI-assisted security research is producing results accepted by Apple. #Apple #Claude #Security #AI ─── 🤖 𝗙𝗼𝗿 𝗺𝗼𝗿𝗲 𝗔𝗜 𝗻𝗲𝘄𝘀 𝗮𝗻𝗱 𝘀𝘁𝗼𝗿𝘆 𝘀𝗼𝘂𝗿𝗰𝗲𝘀, 𝘀𝗲𝗮𝗿𝗰𝗵 "𝗚𝗲𝗻𝗔𝗜𝗦𝗽𝗼𝘁" 𝗼𝗻 𝗧𝗲𝗹𝗲𝗴𝗿𝗮𝗺
Post summary
Apple announced two new CVEs in macOS Tahoe discovered with AI, describing an integer overflow and a WebKit flaw, but no exploits, patches, or PoC were provided.
apple credited claude for finding a kernel vuln in macos 26.5. CVE-2026-28952. like... the AI is writing its own cve entries now?? this is a different chapter
Post summary
Apple credited Claude, an AI model, for discovering a kernel vulnerability (CVE‑2026‑28952) in macOS 26.5, signifying a new chapter in AI involvement in vulnerability reporting.
Also new: MiniMax M2.5 FP4 in the registry (pricing TBD). And @AnthropicAI's Claude independently found CVE-2026-28952 a macOS 26.5 kernel vuln. AI finding zero-days is no longer theoretical.🔬
Post summary
Anthropic AI’s Claude discovered a new macOS 26.5 kernel vulnerability, CVE-2026-28952, and disclosed it. No exploitation details, patches, or PoC information are provided.
The same AI capability that found CVE-2026-28952 is being used inside your org right now, by developers, security engineers, researchers. With a gateway layer, you can see exactly what's being sent to those models before it leaves your perimeter. Visibility first.📊
Post summary
The post highlights an AI tool used internally to detect CVE-2026-28952 and monitor model traffic, but it offers no exploit details, patches, or technical depth.
1/ Claude discovered CVE-2026-28952, a kernel vulnerability in macOS 26.5 now patched by Apple First time an LLM has found a production OS kernel vuln that made it into official security advisories, sets the bar for what automated code review can catch.
Post summary
A new kernel CVE (CVE-2026-28952) was discovered by Claude in macOS 26.5 and has already been patched by Apple, marking the first production OS kernel vulnerability found by an LLM.
Claude found CVE-2026-28952, a macOS 26.5 kernel vuln. Google beats OpenAI's math breakthrough 9 to 1. Pope Leo XIV's AI encyclical drops. 2026-05-26. Thread 🧵
Post summary
The post only notes the existence of CVE-2026-28952 as a macOS 26.5 kernel vulnerability, without further technical detail or actionable information.
Claude found a macOS kernel vulnerability. CVE-2026-28952, credited on Apple's official security page. We debated for years whether AI would be used to attack or defend. The defenders got there first. Security researchers: what's your LLM-assisted workflow right now?
Post summary
Claude reported a new macOS kernel vulnerability (CVE‑2026‑28952) that Apple has documented on its official security page, indicating a recent disclosure with no exploit or patch details provided.
Apple's macOS Tahoe 26.5 patch credits http://Calif.io, Claude, and Anthropic Research for discovering CVE-2026-28952, a kernel privilege escalation vulnerability on M5 Macs — believed to be the first time an AI model has been named in an Apple security bulletin. https://t.co/6M9u5VISUP
Post summary
Apple released macOS Tahoe 26.5 to address CVE‑2026‑28952, a kernel privilege escalation flaw on M5 Macs, crediting AI research for the discovery.
@lostbutlucky Was it CVE-2026-28952?
Post summary
The tweet merely asks if a specific CVE is the relevant vulnerability, with no additional information.
CVE-2026-28952: Apple macOS 26.5 Kernel Vuln found by Claude https://support.apple.com/en-us/127115 (https://news.ycombinator.com/item?id=48273169)
Post summary
Apple identified a kernel vulnerability (CVE‑2026‑28952) in macOS 26.5, with a link to the vendor advisory available.
Apple 安全更新首次致謝 AI。唔係工程師，係 Claude。 CVE-2026-28952 借助 Claude 發現，Apple 安全文件直接致謝 Anthropic。繼逾兩萬漏洞掃出後，AI 協助安全研究正式有官方紀錄。 由「Mythos 太危險」到「Apple 致謝」，幾個月完成反轉。 #AI #ClaudeAI #TechPulseHK https://news.google.com/rss/articles/CBMicEFVX3lxTFBDdlNLWlRBeFM2TFlReGhpUFpnOWNjNE50RDNpcUtSaldEMDlOcHNsM0Itb2RVT0RRTC1MNkljei1YaVhQanhreUVLRkNCUWhkRk9vVm0wc2hMS05DSktqNkpxR3NETEd0MW1PX0JzaFM Score: 72/100 reviewed: false
Post summary
The article reports that CVE‑2026‑28952 was identified with the help of Anthropic’s Claude AI and that Apple’s security documentation now expressly thanks the AI tool, marking the first official acknowledgment of AI assistance in vulnerability discovery.
CVE-2026-28952 : Claude crédité pour une faille noyau APFS de macOS 26.5 corrigée le 11 mai 2026. Une IA nommée dans un advisory Apple. http://lagazetteia.fr/ia-generale/cve-2026-28952-claude-decouvre-une-faille-critique-du-noyau/ #IA #Cybersecurite #macOS
Post summary
Apple announced CVE-2026-28952, a kernel-level APFS vulnerability in macOS 26.5 discovered by the AI Claude, and released a patch on May 11, 2026.
Claude findet kritische macOS-Lücke in Stunden -- Menschen hätten Wochen gebraucht. CVE-2026-28952 ermöglicht komplette Systemkontrolle. KI revolutioniert Cybersecurity. Aber Kriminelle nutzen dieselbe Technologie.
Post summary
AI‑powered tool Claude rapidly identified a critical macOS vulnerability, CVE‑2026‑28952, that potentially permits full system compromise, while noting that criminals are leveraging similar technologies.
CVE-2026-28952: Apple macOS 26.5 Kernel Vuln found by Claude https://support.apple.com/en-us/127115
Post summary
Apple’s support article for CVE-2026-28952 indicates a kernel vulnerability in macOS 26.5 with a patch or mitigation available, but it does not mention PoC, exploits, or active exploitation.
CVE-2026-28952: Apple macOS 26.5 Kernel Vuln found by Claude https://support.apple.com/en-us/127115 (https://news.ycombinator.com/item?id=48273169)
Post summary
A new macOS kernel vulnerability (CVE-2026-28952) was discovered by Claude, and Apple has issued a support advisory that likely contains patch or remediation details.
CVE-2026-28952: Apple macOS 26.5 Kernel Vuln found by Claude https://support.apple.com/en-us/127115 #TechNews #startups
Post summary
The post announces the discovery of CVE-2026-28952, a kernel vulnerability in macOS 26.5, and provides a link to an Apple support page.
Apple's macOS Tahoe 26.5 security notes credit Claude and Anthropic Research for two CVEs: a kernel integer overflow (CVE-2026-28952) and a WebKit use-after-free (CVE-2026-28942). An AI model is now officially credited in Apple's OS security release.
Post summary
Apple’s macOS security notes credit AI research for discovering two CVEs—a kernel integer overflow and a WebKit use‑after‑free—underscoring the growing role of AI in vulnerability identification.
3 of 3 entries
| Part | Vendor | Product | Version | Target SW | Target HW |
|---|---|---|---|---|---|
| OS | apple | ipados | - | - | - |
| OS | apple | iphone_os | - | - | - |
| OS | apple | macos | - | - | - |