🚨 BREAKING: Anthropic’s Most Dangerous Model Ever Breached By Hackers
> anthropic builds a cyberweapon
> calls it mythos
> “can hack every major OS and browser”
> dario: “we’re the safe & responsible ai lab”
> “can’t release it to the public”
> Mercor (their training contractor) gets breached
> leaks anthropic’s model naming conventions
> hackers guess the URL pattern
> contractor credentials still work
> they’re inside
The group also has access to other unreleased Anthropic models.
Not just Mythos. The whole pipeline.
Anthropic’s statement: “investigating a report of access through one of our third-party vendor environments.”
Mythos got breached on day one 💀
> anthropic builds a cyberweapon
> calls it mythos
> “can hack every major OS and browser”
> dario: “we’re the safe & responsible ai lab”
> “can’t release it to the public”
> Mercor (their training contractor) gets breached
> leaks anthropic’s model naming conventions
> hackers guess the URL pattern
> contractor credentials still work
> they’re inside
The group also has access to other unreleased Anthropic models.
Not just Mythos. The whole pipeline.
Anthropic’s statement: “investigating a report of access through one of our third-party vendor environments.”
Mythos got breached on day one 💀
👾46
❗️ Meta has installed trackers on employees' computers and plans to train AI on their mouse movements and keystrokes.
At the same time, Meta is preparing to lay off 10% of its global workforce starting May 20, with more cuts to follow later this year.
At the same time, Meta is preparing to lay off 10% of its global workforce starting May 20, with more cuts to follow later this year.
🤣7
Xiaomi just released MiMo-V2.5-Pro on OpenRouter.
$1 per million input tokens. $3 per million output. 1M context window.
Another Chinese frontier model at a fraction of Opus pricing.
Xiaomi. Moonshot. Alibaba. Z.ai. Four Chinese labs shipping frontier models in a month.
$1 per million input tokens. $3 per million output. 1M context window.
Another Chinese frontier model at a fraction of Opus pricing.
Xiaomi. Moonshot. Alibaba. Z.ai. Four Chinese labs shipping frontier models in a month.
👾6
JUST IN: Google says AI now generates 75% of its new code.
👾4
OPENAI 🚨: WORKSPACE 24/7 AGENTS ARE NOW AVAILABLE ON CHATGPT BUSINESS, ENTERPRISE, AND EDU PLANS.
New ChatGPT Agents are powered by Codex, can use Skills, Connectors, and execute scheduled actions.
👀
New ChatGPT Agents are powered by Codex, can use Skills, Connectors, and execute scheduled actions.
👀
🔥3
NEW: Chinese cybersecurity firm 360 Digital unveils new AI “Vulnerability Discovery Agent” that seeks to rival Claude Mythos.
❤3
Mozilla has patched 271 vulnerabilities in Firefox after using Anthropic’s Claude Mythos Preview AI model to scan its codebase.
Part of Anthropic’s cybersecurity-focused Project Glasswing, the AI identified these issues with performance that matched human-level capability across different vulnerability types.
This follows an earlier scan in which Claude Opus 4.6 discovered 22 sensitive security flaws in Firefox 148, including 14 of high severity.
Mozilla’s CTO described the development as transformative for defensive security work on large open-source projects.
👾44
Open Ai
GPT-Image-2 by @OpenAI has claimed the #1 spot across all Image Arena leaderboards!
GPT-Image-2 had a 93% win rate in Image Arena.
JUST IN: 🚨🤖 Anthropic says it has no way to control or shut down its AI models once deployed by the Pentagon, according to a new court filing.
👾10💩3
🇨🇳🇺🇸 US has accused China of "industrial-scale" theft of AI and warned that it will be cracking down on the exploitation, per FT.
👾43🤣1
🚨 NEW: Sam Altman alleges Anthropic is using “fear-based marketing” for Claude Mythos and may be overstating AI risks.
👾44
JUST IN: Meta announces it's laying off 10% of its workforce, as the company plans on AI writing 4x the amount of code as its human engineers this year.
#layoff
#layoff
👾44
🚨Openai launched ChatGpt 5.5🤖
> self-verifying outputs.
> same speed, smarter.
> persists across tools until the task is done.
> uses fewer tokens than 5.4 for the same work
> beats Claude Opus 4.7 and Gemini 3.1 Pro on agentic benchmarks.
OpenAI didn’t hold back with this one.
Please open Telegram to view this post
VIEW IN TELEGRAM
👾46