The Financial Times first reported the specialized defense partnership.
Danielle Fong 🔆#756
Interesting that it's confirmed, although I basically assumed this was happening.
SCOOP #NSA is using #Mythos to conduct offensive cyber operations. Anthropic engineers are embedded in the US intelligence agency.
(@CristinaCriddle & me @AsiaLens)
https://www.ft.com/content/d02d91b3-2636-454e-9442-dc7e69f51815
Many users criticized Anthropic's collaboration with the NSA to deploy Mythos for offensive cyber operations as hypocritical virtue signaling that undermines their AI safety and ethics claims.
waow
Sources: Anthropic has embedded around half a dozen forward-deployed engineers within the NSA to help the agency deploy Mythos for offensive cyber operations (Financial Times)
(Visit Techmeme dot com for the link and full context!)
SuPpLy ChAiN rIsK
SCOOP #NSA is using #Mythos to conduct offensive cyber operations. Anthropic engineers are embedded in the US intelligence agency.
(@CristinaCriddle & me @AsiaLens)
https://www.ft.com/content/d02d91b3-2636-454e-9442-dc7e69f51815
So autonomous weapons aren't okay but offensive cyberwarfare for the NSA is?
Financial Times: "one person close to the situation said Mythos would be useful for infiltrating the networks of nations such as China or Iran"
So autonomous weapons aren't okay but offensive cyberwarfare for the NSA is?
I think very soon a big divergence between OpenAI&Anthropic and everyone else becomes visible. Maybe Google wakes up too. From there and until Q1 2027 at least, we will be seeing burgers at their worst. Smug, showboating, basking in the limelight, reckless. Assured of impunity.
soon… Actually, NOW.
soon… Actually, NOW.
So autonomous weapons aren't okay but offensive cyberwarfare for the NSA is?
okie dokie
Sources: Anthropic has embedded around half a dozen forward-deployed engineers within the NSA to help the agency deploy Mythos for offensive cyber operations (Financial Times)
(Visit Techmeme dot com for the link and full context!)
🤔
@teortaxesTex The fire rises.
soon… Actually, NOW.
@deanwball completely and permanently beclowning their moral authority to be able to regulate this. the most conflicted interest imaginable
SuPpLy ChAiN rIsK
https://www.ft.com/content/d02d91b3-2636-454e-9442-dc7e69f51815?shareType=nongift
SCOOP #NSA is using #Mythos to conduct offensive cyber operations. Anthropic engineers are embedded in the US intelligence agency.
(@CristinaCriddle & me @AsiaLens)
https://www.ft.com/content/d02d91b3-2636-454e-9442-dc7e69f51815
@scaling01 Yes, seems fine.
Remember that Anthropic's objection to autonomous AI weapons was that the current models aren't good enough for this use case.
https://www.ft.com/content/d02d91b3-2636-454e-9442-dc7e69f51815
@AndrewCurran_ tbh no idea what it matters is it a good SMR?
@teortaxesTex The fire rises.
@deanwball
Anthropic is conducting “offensive cyber operations” with the NSA
Alignment huh?
FT Exclusive: Anthropic is helping the US National Security Agency deploy its Mythos AI model for offensive cyber operations despite a legal battle with the Pentagon. https://ft.trib.al/65Qvavy
@Techmeme claude went from "i can't help with that" to nsa contractor real quick
@CristinaCriddle @AsiaLens Mythos and other "cyber-capable" models need to be secure themselves! https://github.com/OraclesTech/guardian-sdk
If I were anthropic the above tweet/qt pair would constitute the entirety of my defense in court
