brianmcnulty

5 ms·

1.
by brianmcnulty 3mo ago
Long time no see! It's been a while since I've looked at MCP, but I think this does a really good job at making MCP more secure for organizations and addressing some of the weaknesses of dynamic client registration. Now that clien
2.
by brianmcnulty 3mo ago
Why would you have ethics when you could get that IPO money instead?
3.
by brianmcnulty 3mo ago
I wonder how Claude Fable will live up to expectations and how good those Fable/Mythos classifiers really are. It seems a bit convenient for Anthropic to release this magical insane model when they are about to IPO.
4.
by brianmcnulty 3mo ago
That seems to conflict with the recent security blog that says they are using Google Cloud infra and NVIDIA GPUs with PCC now [0]. They are allowing it to run on Intel and NVIDIA and Google chips meeting certain requirements now too instead
5.
by brianmcnulty 3mo ago
They do this by allowing you to download all of the components (minus data cryptexes containing the model weights) and run it on your own Apple silicon chip (you can put your computer in recovery mode and use csrutil to enable research gues
6.
by brianmcnulty 3mo ago
No, it's more that those apps needs to be able to make all of the tool calls Siri AI can make, which would allow third-party developers to collect data they shouldn't have access to. App developers can already access the on-device
7.
by brianmcnulty 3mo ago
I think it's because Apple would have to provide every competitor (including ones running off-device with no confidential compute) with the same level of access Siri AI would get, which poses a lot of security and privacy concerns Appl
8.
Microsoft Scout: Your always-on personal agent(microsoft.com)
2 points by brianmcnulty 4mo ago|1 comments
9.
by brianmcnulty 4mo ago
Yeah, it seems another ATO bug has popped up. I haven’t looked too much at it personally, but I hope Meta plans on taking their Meta AI Support Assistant offline until it undergoes far more rigorous security review.
10.
by brianmcnulty 4mo ago
Not exactly, I think part of the tools it has access to allows it to perform an "investigation" into recent malicious activity and account changes that may have occurred that were likely unauthorized (such as changing from an emai
11.
by brianmcnulty 4mo ago
You should also assume the user can read any data you send back from a tool call or data you add to a user response. If any part of the input or output is controllable by an attacker, you should be assuming some prompt injection is possible
12.
by brianmcnulty 4mo ago
I do a lot of bug bounty research on Meta and Instagram, and some of the bugs I find look extremely simple like this but have some slightly complicated reason for why they occur. Maybe not this one, but I do have a guess as to what might ha
13.
Investigation update: GitHub Enterprise Server signing key rotation(github.blog)
2 points by brianmcnulty 4mo ago|0 comments
14.
Staged publishing and new install-time controls for npm(github.blog)
61 points by brianmcnulty 4mo ago|11 comments
15.
by brianmcnulty 4mo ago
Based on what I can tell, this bug just allows a persistent service worker to run forever by downloading a large file and not letting it complete? Security impact is pretty limited (but definitely not none). It can make requests but only wi
16.
Meta AI introduces private confidential-compute backed Incognito Chats(about.fb.com)
4 points by brianmcnulty 4mo ago|0 comments
17.
Claude Code RCE: Exploiting Deeplink Handlers via Settings Injection(0day.click)
2 points by brianmcnulty 4mo ago|0 comments
18.
by brianmcnulty 5mo ago
I assume a fair amount of these on-prem customers restrict access to their GHES instance to be behind corporate VPN or something similar and are planning a date to upgrade their instance that won't affect operations. Any public instanc
19.
by brianmcnulty 5mo ago
The tweet is confusing and makes it sound like the RCE was as simple as `git push -o "x;`whatever command`"`, but there are a few more things they have to specify that they mention in their blog post: https://www.wiz.io
20.
by brianmcnulty 5mo ago
It's interesting that Next is becoming so popular when LLMs supposedly have a capability to work with all these other frameworks that don't create a dependency on something like Vercel.
21.
by brianmcnulty 5mo ago
I heard that JWTs are 5x the price of JSON tokens but only 3x if you have JSON ForULTRA+ (new) (for work or school).
22.
by brianmcnulty 5mo ago
It has been up and down today, specifically with authentication breaking. I also saw an error message with backend SQL in it (in my 6 years of Meta bug bounty security research, I have never once seen backend SQL before). I suspect it is be
23.
by brianmcnulty 6mo ago
I guess this means the listener for Hey Siri requests has to be inside of the exclave/conclave to avoid triggering the mic indicator light 24/7 or leaking microphone data? I assume this means the code has to be able to be updated