Anthropic has reportedly finished training a more capable successor.
NSA SHOCKER: MYTHOS BREACHED NEARLY ALL CLASSIFIED SYSTEMS IN HOURS, NOT WEEKS
NSA CONFIRMS MYTHOS PENETRATED MOST CLASSIFIED NETWORKS WITHIN HOURS
CYBER EARTHQUAKE: MYTHOS CRACKED TOP U.S. SECRET SYSTEMS IN HOURS
Positive users see Mythos AI's rapid NSA network breach as beneficial for targeting adversaries like the IRGC or proof of strong models, while negative users dismiss the reports as fake news, marketing hype, or evidence of NSA incompetence.
No Digg Deeper questions have been answered for this story yet.
@FirstSquawk I call BS because the most classified networks are air gapped.
Humans orchestrated the testing of Anthropic’s Mythos AI model (which NSA has been using for cyber ops).
During red-teaming, Mythos autonomously found and chained exploits to penetrate nearly all classified systems in hours — not weeks.
Humans set the task and prompt; the AI executed at machine speed with high independence. Not a rogue breach, but a stark demo of frontier AI cyber capability.
@FirstSquawk Guess “classified” now just means “Mythos already read it.” 😂
@FirstSquawk A more accurate headline: "NSA uses Mythos to test their system security".
@FirstSquawk Why don't they use Mythos against Iran?
So Mythos "broke into almost all of NSA's classified systems, not in weeks, but in hours" and we already got the next iteration of Mythos.
Big things are happening.
Holy Sh*t: that changes the whole Fable 5 story completely:
On June 11, the very same day Amazon reportedly uncovered the jailbreak, “Mythos” allegedly breached almost all classified systems belonging to the NSA and U.S. Cyber Command, not over the course of weeks, but within hours.
"On June 11th Mark Warner, the vice-chair of the Senate Intelligence Committee, said that General Joshua Rudd, who leads the National Security Agency and the Pentagon’s Cyber Command, had told him that Mythos “broke into almost all of our classified systems, not in weeks, but in hours”."
Via Economist
@FirstSquawk @Liuskaluiska Skynet was the prediction
@FirstSquawk @grok is this humans orchestrating the breach or did it do it by itself?
@FirstSquawk Can it get us a copy of the Epstein files while it's in there?
@traders_guild @FirstSquawk The important ones are Air Gapped.
@Triode_in_situ @FirstSquawk yes hard to believe all these claims, maybe its to pump the ipo
@FirstSquawk This is a condemnation of how fucking retarded the nsa is. Not how competent AI is.
@FirstSquawk Highly doubt they would “announce” it like that.
@FirstSquawk Mythos "Breaching" almost all of the systems of the NSA. This is far from the truth. In reality, it's much different. More details about it here (what's it and what's it not): https://thecybersecguru.com/news/mythos-nsa-breach-claim/
@FirstSquawk They are trying to justify keeping AI for gov’t use only.
@FirstSquawk Show me a piece of data that is not compromised.
@FirstSquawk Hmm where is the source for this? Also at what level?
Good grief!
“• The core claim is accurate: per The Economist, Senator Mark Warner reported that NSA/Cyber Command head General Joshua Rudd stated Anthropic’s Mythos AI “broke into almost all of our classified systems, not in weeks, but in hours.” • Mythos is Anthropic’s restricted, high-capability AI model optimized for offensive cyber tasks like chaining exploits and identifying vulnerabilities; the NSA has used it despite DoD supply-chain concerns.”
@FirstSquawk It's looking like the best security is how it has been throughout history: IN-PERSON MEETINGS, AND AIR-GAPPED PHYSICAL DEVICES THAT ONLY CONNECT IF YOU PHYSICALLY CONNECT THEM.
@Equalizetheset @FirstSquawk Palantir was used in Iran. It could bring up 100 potential military sites to attack in seconds.
This is why the schoolkids and school building were destroyed. It's very effective...
