alt.hn

6/4/2026 at 8:33:02 PM

NSA using Anthropic's Mythos for cyber attacks

 https://www.ft.com/content/d02d91b3-2636-454e-9442-dc7e69f51815

by jawiggins

6/4/2026 at 9:59:25 PM

The FT article has a lack of evidence of an actual operation. It claims Mythos is being used for "offensive cyber operations" in the first paragraph, but doesn't give a source supporting this claim. The lone piece of evidence given says Mythos "would be useful for" infiltrating China or Iran, which is a hypothetical, not evidence of an actual operation.

None of that means the NSA isn't doing offensive work, my issue is that FT claims Mythos is involved in active foreign attacks based on a single anonymous source saying it "would be useful."

by tristanj

6/4/2026 at 10:49:02 PM

Right, and zero non-anonymous sources. A person plausibly "close to Anthropic" or vaguely "with knowledge of the situation" using the stock phrase "the best defense is a good offense"... this is not a news article y'all. It's speculation.

by _alternator_

6/5/2026 at 11:40:02 PM

Meaning "an LLM related to"

by LearnYouALisp

6/4/2026 at 11:09:34 PM

speculation, or marketing.

by saidnooneever

6/4/2026 at 9:52:10 PM

I am all about NSA using this stuff as "Defense" but my main concern is their long track record of unconstitutional warrantless, or abuse of shadow courts, to surveil the people they're supposed to be protecting, and for nefarious purposes like spying on ex girlfriends [1]

[1] - https://www.reuters.com/article/world/uk/nsa-staff-used-spy-...

by maerF0x0

6/5/2026 at 1:09:02 AM

And most of that happened under comparatively good administrations. Imagine what’s been greenlit now.

by ninjalanternshk

6/4/2026 at 10:59:21 PM

Wait til you hear about other countries, to say nothing of Silicon Valley.

by applicative

6/5/2026 at 3:25:17 PM

Two wrongs don't make a right

by maerF0x0

6/4/2026 at 10:04:13 PM

dont worry, theyll also outspurce to capitalists to ensure all the paychopaths get a taste

by cyanydeez

6/4/2026 at 10:25:57 PM

You just typo’d the words “outspurce” and “paychopaths” into existence =D

Outspurce - to pay someone to do your (software) dirty work

Paychopath - someone who goes against their ethics as soon as the price is right

Two in one post. New record!

by cadamsdotcom

6/4/2026 at 9:48:34 PM

while this article is frustratingly light on substance, it should be no surprise the nsa would use anything and everything that could give them an edge. of course they would test out mythos, chatgpt-cyber, etc.

by john_strinlai

6/4/2026 at 10:40:12 PM

accurate and they probably had access already.

by selectedambient

6/4/2026 at 9:30:08 PM

That headline is doing a lot of heavy making-it-up summarization.

by thrill

6/4/2026 at 9:35:43 PM

Yeah, I honestly feel like some sources should be banned if they regularly editorialize titles like this. FT wants to sell subs by any means necessary, even if it means creating some clearly misleading titles to do so. By amplifying their reach and allowing posts like that only further incentivizes them to continue to do it.

by NewsaHackO

6/4/2026 at 9:30:12 PM

""" It remains unclear whether Anthropic’s engineers are assisting the NSA in active operations. However, one person close to the situation said Mythos would be useful for infiltrating the networks of nations such as China or Iran. """

by diwank

6/4/2026 at 9:42:24 PM

seems they forgot to say "...and the networks of friendly nations / allies"

by keithnz

6/5/2026 at 6:04:32 AM

Mythos useful for infiltrating networks of China and Iran?

Anyone who’s read anything about Mythos would’ve said the same thing on day one. Didn’t need some anonymous vague source to tell you that.

by atonse

6/4/2026 at 11:17:22 PM

This feels like a stupid idea, but it's not my tax money being wasted on it (yet).

An AI being used for offensive purposes would be so easy to counter due to Brandolini's Law. A much simpler AI could probably easily create dumb honey pot systems that looked like real systems with holes in them to lead more advanced and expensive systems on a wild goose chase.

Suddenly an AI creating security vulnerabilities is a core part of a future cyber defence strategy. Surely someone must be doing that already

by LarsKrimi

6/4/2026 at 9:56:32 PM

Not unexpected. Is anyone tracking which episode we're on in the Pantheon timeline?

by sigmar

6/4/2026 at 9:48:29 PM

The only relevant except: "The San Francisco-based company had installed about half a dozen staff within the NSA as so-called forward-deployed engineers to guide the use of the technology and customise models for specific applications, two people familiar with the arrangement said.

It remains unclear whether Anthropic’s engineers are assisting the NSA in active operations. However, one person close to the situation said Mythos would be useful for infiltrating the networks of nations such as China or Iran.

“The best way to build a good defence is to build a good attack,” said a person close to Anthropic, who argued that adversaries are probably building their own AI-driven offensive technology. “If [Mythos] is not used to build attack agents, adversaries will find a way to do it.”"

by Computer0

6/5/2026 at 5:36:53 AM

The use of AI is to be expected. NSA's bigger evils are them sabotaging our use of good encryption, of making the internet architecturally weak, and forcing the hand of companies to hand over user data to the government.

by OutOfHere

6/5/2026 at 12:54:44 AM

But HN told me it’s just a marketing stunt

by reducesuffering

6/5/2026 at 12:23:27 AM

I'd like to see Ukraine using it against Russia, if they're not doing it already.

by aussieguy1234

6/4/2026 at 9:25:10 PM

How to get past this paywall ?

by maerF0x0

6/4/2026 at 9:27:28 PM

Use mythos to bypass.

by NuclearPM

6/4/2026 at 9:46:49 PM

With a credit card ;-)

by ale42