1、 当地时间3月27日，美国和以色列对伊朗发动的军事打击进入第28天， 美以对伊朗多处工业及核相关设施发 动空袭。 伊朗则以无人机和导弹还击，并警告报复美以关联的工业企业。 伊朗伊斯兰革命卫队发表声明称，霍 尔木兹海峡目前已经关闭， 任何试图通过该海峡的行为都将遭到严厉打击。 伊朗官员称考虑退出《不扩散核武 条约》；一名伊朗高级安全官员警告，若美国发动地面行动， 将遭对等反制。 2、 美国国防部官员称，五角大楼正在考虑向中东增派多达1万人的地面部队，以便在美国总统特朗普权衡与伊 朗和平谈判的同时， 为其提供更多军事选择。 以色列国防部长卡茨表示，将进一步升级对伊朗的打击行动，扩 大打击范围，覆盖更多目标和区域。也门胡塞武装表示，将毫不犹豫地履行其责任和义务，支持伊朗，对抗以 色列及其"美国帮凶"。 3、 美国总统特使威特科夫表示， 预计本周将与伊朗方面举行会谈。美国副总统万斯在本周早些时候一次电话 通话中"批评以色列总理内塔尼亚胡对本次美以伊冲突的预测过于乐观"，认为内塔尼亚胡"夸大了伊朗政权更迭 • 国务院常务会议要求扩大服务业市场准入和开放领域 • 商务部：对美国发起两项贸易壁垒调查 • 今年前两个月我 ...

for market on close. I'm Marley Hayden here in Chicago alongside Sam Botis at the New York Stock Exchange. We'll close out the week with our final thoughts on the session.Today we saw cyber security stocks slumping on a really interesting report that Anthropic is testing a powerful new AI model called Claude Mythos. Now, Fortune First reporting this news, citing details from a publicly accessible draft of a blog post that included nearly 3,000 draft documents about the model's advanced coding and cyber secu ...

CrowdStrike Holdings Inc (NASDAQ: CRWD) is inching lower on Mar 27 after partner Anthropic previewed a "breakthrough†AI model with advanced cybersecurity capabilities it's calling Claude Mythos. Adding to the pressure are reports of Amazon.com Inc, another one of CRWD's partners, developing a sophisticated artificial intelligence tool to handle sales, business development, and other tasks. ...

Fortinet stock is taking a hit today. What’s behind FTNT decline?Anthropic Leak And AI Security Fears Hit Cyber NamesCybersecurity stocks moved lower after Anthropic allegedly exposed internal data through a content management system misconfiguration, with the lapse reportedly revealing details about an unreleased model called Claude Mythos. Anthropic later said the issue was caused by human error and was unrelated to its AI tools, but the event still added fresh anxiety around how quickly AI is changing th ...

Cybersecurity Stocks Take a Hit: Palo Alto Drops 6%, Okta Tumbles 7% on AI Competition Fears - 24/7 Wall St. S&P 5006,420.90 -1.21% Dow Jones45,542.20 -1.22% Nasdaq 10023,321.80 -1.42% Russell 20002,465.66 -1.59% FTSE 1009,989.60 +0.30% Nikkei 22551,929.50 -1.43% Stock Market Live March 27, 2026: S&P 500 (SPY) Slips Despite Trump Deadline Extension Investing Cybersecurity Stocks Take a Hit: Palo Alto Drops 6%, Okta Tumbles 7% on AI Competition Fears By David MoadelPublished Mar 27, 1:00PM EDT Quick Read ...

Shares of Tenable Holdings (NASDAQ: TENB), a cybersecurity company, were falling hard this morning after a leaked Anthropic blog post claimed that a yet-to-be-released Claude model has advanced capabilities that could surpass current cybersecurity capabilities. That is, of course, a bit unsettling for any company trying to protect data, but it's especially worrying for Tenable and other cybersecurity stocks. Tenable stock plunged on the news and is down 8.5% as of 11:31 a.m. ET. Will AI create the world's ...

Palo Alto Networks (NASDAQ: PANW) stock, one of the nation's biggest cybersecurity companies, fell an unlucky 7% through 9:55 a.m. ET Friday. What's up with Palo Alto going down? It seems artificial intelligence company Anthropic has a new model coming out soon, "Claude Mythos," and it's bad news for cyber companies like Palo Alto. Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technol ...

About this content About Emily Jarvie Emily began her career as a political journalist for Australian Community Media in Hobart, Tasmania. After she relocated to Toronto, Canada, she reported on business, legal, and scientific developments in the emerging psychedelics sector before joining Proactive in 2022. She brings a strong journalism background with her work featured in newspapers, magazines, and digital publications across Australia, Europe, and North America, including The Examiner, The Advocate, ...

作者 ｜ 木子 Anthropic"手滑"泄露了最强新模型，但更戏剧的是，他们本来还不太敢发。 因为这个模型发现漏洞、攻击漏洞的"黑客能力"太强了。 据《财富》消息，Anthropic 在未公开的草稿中指出： "与此前最强模型 Claude Opus 4.6 相比，Capybara（卡皮巴拉，新模型的内部代号）在软 件编程、学术推理和网络安全等测试中取得了显著更高的分数。" 这个"卡皮巴拉"其实还有个真名，叫 "Claude Mythos" （直译过来就是"神话"）。 目前，Anthropic 的 Claude 模型分为三档： 而"神话"Mythos，"模"如其名，确实也是一个比 Opus 更强、更昂贵的新层级。 从表面上看，Mythos 在 编程、推理，以及网络安全测试 中的表现提升，像是三项独立能力的增 强。但在安全领域，这三者实际上恰恰能构成一条完整的"攻击能力链"： 首先，编程能力决定模型能不能真正"看懂系统"：不仅是读代码，还包括理解模块之间如何交互、逻 辑是怎么实现的；而且还可以在需要时，把这些理解转化为可执行的利用代码（exploit）或自动化脚 本。 其次，强悍的推理能力可以做"攻击路径的规划 ...