- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
Jerome Powell and Scott Bessent met with banks to discuss cyber threats from Anthropic
Posted on 4/10/26 at 12:44 pm
Posted on 4/10/26 at 12:44 pm
quote:
Federal Reserve Chairman Jerome Powell and Treasury Secretary Scott Bessent met with major U.S. bank CEOs this week to discuss the possible cyber risks raised by Anthropic’s Mythos model, CNBC confirmed Friday.
The bank heads were already in Washington, D.C., for a Financial Services Forum board meeting when a special gathering was called on Tuesday to discuss Mythos, according to people familiar with the matter, who asked not to be named in order to share information about a confidential matter.
LINK
I think we'll hear a lot more about cybersecurity and AI in the future
6
Posted on 4/10/26 at 12:48 pm to Powerman
quote:
Today we’re announcing Project Glasswing1, a new initiative that brings together Amazon Web Services, Anthropic, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks in an effort to secure the world’s most critical software.
We formed Project Glasswing because of capabilities we’ve observed in a new frontier model trained by Anthropic that we believe could reshape cybersecurity. Claude Mythos2 Preview is a general-purpose, unreleased frontier model that reveals a stark fact: AI models have reached a level of coding capability where they can surpass all but the most skilled humans at finding and exploiting software vulnerabilities.
Anthropic Statement
Posted on 4/10/26 at 1:01 pm to Powerman
quote:
Mythos Preview found a 27-year-old vulnerability in OpenBSD—which has a reputation as one of the most security-hardened operating systems in the world and is used to run firewalls and other critical infrastructure. The vulnerability allowed an attacker to remotely crash any machine running the operating system just by connecting to it;
It also discovered a 16-year-old vulnerability in FFmpeg—which is used by innumerable pieces of software to encode and decode video—in a line of code that automated testing tools had hit five million times without ever catching the problem;
The model autonomously found and chained together several vulnerabilities in the Linux kernel—the software that runs most of the world’s servers—to allow an attacker to escalate from ordinary user access to complete control of the machine.
We might be cooked
Posted on 4/10/26 at 1:05 pm to Powerman
quote:
We might be cooked
We absolutely are
Posted on 4/10/26 at 1:06 pm to HailToTheChiz
This is probably the biggest news of the day
At least we beat the Chinese to it
At least we beat the Chinese to it
Posted on 4/10/26 at 1:10 pm to HailToTheChiz
We are absolutely cooked. Yet we will plow forward with our AI overlords
Posted on 4/10/26 at 1:12 pm to Cosmo
At least anthropic is being very proactive about trying to find new ways to defend against the monsters we're creating
Posted on 4/10/26 at 1:17 pm to Powerman
I’m glad they’re being proactive as long as it’s legitimate concern and not fear mongering. Anthropic’s CEO issued a similar warning when he was at OpenAI for a prior GPT model.
I can very easily see where Claude Code and other similar environments could have the potential to perform the equivalent of a brute force attack on software. The speed at which it can develop and adapt is pretty amazing.
I can very easily see where Claude Code and other similar environments could have the potential to perform the equivalent of a brute force attack on software. The speed at which it can develop and adapt is pretty amazing.
Posted on 4/10/26 at 1:19 pm to Cosmo
I'm just now remembering that Anthropic accidentally posted their source code online
Damn it, I hope it (the leak) didn't have anything to do with that new model on there because you know all sorts of bad actors have that code now
Damn it, I hope it (the leak) didn't have anything to do with that new model on there because you know all sorts of bad actors have that code now
Posted on 4/10/26 at 1:20 pm to TerryDawg03
quote:
as long as it’s legitimate concern and not fear mongering
If they're finding exploits that went unnoticed for over 27 years in some of the most secure software systems in the world...that's huge systemic risk.
Posted on 4/10/26 at 1:21 pm to Powerman
Mattresses and coffee cans now on sale !!!
Posted on 4/10/26 at 1:34 pm to Powerman
There's a difference between finding a vulnerability, and it being easily exploitable.
That type of thing matters.
That type of thing matters.
Posted on 4/10/26 at 1:36 pm to Centinel
quote:
There's a difference between finding a vulnerability, and it being easily exploitable.
That type of thing matters.
Of course
And from my quoted text
quote:
The vulnerability allowed an attacker to remotely crash any machine running the operating system just by connecting to it
I'd call that exploiting it
It was enough of a concern for 2 very important people to meet with US Banks over in an emergency type meeting
Posted on 4/10/26 at 1:39 pm to Powerman
quote:
I'd call that exploiting it
The NSA can exploit all kinds of fun things. 99% of the rest of the people on this planet can't.
That's my point. And it's why folks in my profession are certainly paying attention, but they're not hyperventilating over it.
Risk assessment is a thing.
Posted on 4/10/26 at 1:42 pm to Centinel
quote:
The NSA can exploit all kinds of fun things. 99% of the rest of the people on this planet can't.
That's my point. And it's why folks in my profession are certainly paying attention, but they're not hyperventilating over it.
Risk assessment is a thing.
I think that's sort of at the core of the issue here. They aren't releasing the model because an "operator" type of person could theoretically use it to exploit systems. If they're deeming it too dangerous to release to the public it's worth noting. Powell and Bessent aren't meeting with banks about this because it's trivial.
Posted on 4/10/26 at 1:47 pm to Powerman
quote:
hey aren't releasing the model because an "operator" type of person could theoretically use it to exploit systems.
You're assuming this didn't exist already as a zero-day in one of our group's back pockets, or China's, or Russia's, etc.
Let me try this another way:
What anthropic's model is doing is exposing vulns quicker than they normally would. That's a *good* thing. Because of what I mentioned above. There are all kinds of zero-day exploits out there, in the back pocket of the big boys, that haven't been publicly discovered/acknowledged.
That is what should scare the shite out of you. Not what anthropic's new toy is finding.
You'll just have to trust me on this one. Or not. Macht nichts.
Posted on 4/10/26 at 1:49 pm to Centinel
quote:
What anthropic's model is doing is exposing vulns quicker than they normally would. That's a *good* thing. Because of what I mentioned above. There are all kinds of zero-day exploits out there, in the back pocket of the big boys, that haven't been publicly discovered/acknowledged.
That is what should scare the shite out of you. Not what anthropic's new toy is finding.
Oh I definitely think it's good that they can find these vulnerabilities before bad actors can
This could ultimately be a way to shore up cyber security
Posted on 4/10/26 at 1:50 pm to Powerman
quote:
This could ultimately be a way to shore up cyber security
Oh, it certainly is. My current employer is jizzing themselves over it right now.
ETA: Oh, and there are a bunch of boys in three letter agencies that are NOT happy right now.
Well, assuming they're not already one step ahead. Which they probably are. I've been out of that game for a few years now.
This post was edited on 4/10/26 at 1:53 pm
Posted on 4/10/26 at 1:51 pm to Centinel
I know crowdstrike chimed in and said it's a game changer
So that's cause for optimism
So that's cause for optimism
Posted on 4/10/26 at 1:52 pm to Centinel
They have brought in all the major SW players to use this to help harden their systems.
This is a big deal and something that could become a strategic weapon if the NSA/CIA/DOD wanted to weaponize this. Anthropic spurned their advances to use their AI for military purposes ( Taft Law Article) and they banned all federal agencies from using Anthropic tools
This is a big deal and something that could become a strategic weapon if the NSA/CIA/DOD wanted to weaponize this. Anthropic spurned their advances to use their AI for military purposes ( Taft Law Article) and they banned all federal agencies from using Anthropic tools
Page 1 of 2
Popular
Back to top
Follow TigerDroppings for LSU Football News