Anthropic的Mythos AI发现加密漏洞，引发安全质疑 · 跟踪10个来源

报道来源 [10]

1. LessWrong (AI tag) TIER_1 English(EN) · Philipreal · 2026-07-29 16:21

   关于 Anthropic 加密博客文章的笔记

   <p><i><span>Status: Mostly a summary with some of my notes</span></i><span> </span><i><span>at the end. </span></i></p><p><br /></p><p><span>Anthropic released </span><a href="https://www.anthropic.com/research/discovering-cryptographic-weaknesses"><span>a blogpost</span></a><spa…

2. The Decoder TIER_1 English(EN) · Matthias Bastian · 2026-07-28 19:12

   Anthropic表示其Mythos模型发现了保护互联网的加密算法中的漏洞

   <p><img alt="" class="attachment-full size-full wp-post-image" height="768" src="https://the-decoder.com/wp-content/uploads/2026/07/anthropic_cybersecurity.png" style="height: auto; margin-bottom: 10px;" width="1376" /></p> <p> Anthropic's Claude Mythos Preview found weaknesses i…

3. HN — anthropic stories TIER_1 English(EN) · supermatou · 2026-07-29 16:42

   关于Anthropic新密码分析结果的一些想法

4. Mastodon — sigmoid.social TIER_1 English(EN) · [email protected] · 2026-07-29 14:05

   Anthropic的Claude Mythos预览通过识别加密算法中的核心数学漏洞，展示了重大突破，包括

   Anthropic's Claude Mythos Preview has demonstrated a significant breakthrough by identifying core mathematical vulnerabilities in cryptographic algorithms, including HAWK and round-reduced AES. In response to these findings, Anthropic is collaborating with NATO and major technolo…

   链接 anthropic.com/…/discovering-cryptographic…
5. Mastodon — fosstodon.org TIER_1 Deutsch(DE) · [email protected] · 2026-07-31 08:00

   Anthropic的Claude在加密算法中发现漏洞 | 安全

   Anthropics Claude entdeckt Schwachstelle in Krypto-Algorithmus | Security https://www. heise.de/news/Anthropics-Claud e-entdeckt-Schwachstelle-in-Krypto-Algorithmus-11384965.html # ArtificialIntelligence # AI # Anthropic # AnthropicMythos # AnthropicClaude # AnthropicClaudeMythos…

   链接 heise.de/…/Anthropics-Claude-entdeckt-Sch…
6. Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] · 2026-07-29 21:03

   Anthropic 的 Mythos 发现了抗量子加密算法 HAWK 和 AES 的弱点。目前没有直接风险，但这凸显了前沿人工智能

   Anthropic’s Mythos found weaknesses in quantum resistant encryption algorithms HAWK and AES. There is no immediate risk, but this it highlights how frontier AI models are significantly increasing their capabilities to identify and exploit weaknesses in software, hardware, algorit…

   链接 matthewrosenquist.substack.com/…/mythos-f… matthewr.substack.com
7. Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] · 2026-07-28 23:30

   Anthropic 的 Mythos 模型在 60 小时内以约 10 万美元的成本识别出后量子加密方案 HAWK 的漏洞，研究结果不影响当前

   Anthropic's Mythos model identified vulnerabilities in HAWK, a post-quantum cryptographic scheme, in 60 hours at a cost of ~$100K. Findings don’t impact current systems but question long-term security assumptions. Source: The Decoder AI https:// the-decoder.com/anthropic-says -it…

   链接 the-decoder.com/anthropic-says-its-mythos…
8. Mastodon — mastodon.social TIER_1 English(EN) · [email protected] · 2026-07-29 16:01

   🤖 Mythos 首次尝试后量子密码学 📝 Anthropic 的 Claude Mythos Preview 模型已帮助研究人员发现加速攻击的方法...

   🤖 Mythos takes its first shot at post-quantum cryptography 📝 Anthropic’s Claude Mythos Preview model has helped researchers discover ways to speed up attac... https://www. csoonline.com/article/4202920/ mythos-takes-its-first-shot-at-post-quantum-cryptography.html 📰 CSO Online # …

   链接 csoonline.com/…/mythos-takes-its-first-sh… csoonline.com/…/4202920
9. Mastodon — mastodon.social TIER_1 Deutsch(DE) · aisyndicate · 2026-07-29 06:10

   Anthropic的Mythos预览识别出HAWK等加密算法中的漏洞。使用60小时API时间显示实际计算负载

   Anthropic's Mythos Preview identifiziert Schwachstellen in kryptografischen Algorithmen wie HAWK. Der Einsatz von 60 Stunden API-Zeit zeigt die reale Rechenlast für automatische Sicherheitsaudits. https:// the-decoder.de/anthropics-ki-m odell-mythos-preview-findet-schwachstellen-…

   链接 the-decoder.de/anthropics-ki-modell-mytho…
10. Mastodon — mastodon.social TIER_1 Français(FR) · [email protected] · 2026-07-28 22:00

    Anthropic 发布其 Mythos 模型，据称在后量子加密方案中发现研究人员的漏洞

    Anthropic a lâché son modèle Mythos sur des algorithmes de chiffrement — et il aurait trouvé des failles dans des schémas post-quantiques sur lesquels des chercheurs humains butaient depuis des années. À noter : "cracked" mérite d'être lu avec soin. S'agit-il d'attaques pratiques…

    链接 cyberscoop.com/anthropic-claude-mythos-en… cyberscoop.com/anthropic-claude-breaks-ba…