Had a fun time figuring this one out today! ghost.ai.moda/google-gemini-…

tfw so 1337 ChatGPT credits me in fake exploits

It's official, ai.moda will be be competing in DARPA's AIxCC semifinals!

Let's see if OpenAI pays..

Sometimes I have a hard time understanding mitigations.. ..because most of these mitigations have never broken a PoC of mine.

Worth it?

😬 that's a hard miss Gemini.

Great find of a vulnerability in well defended software, but also a great write up of the find, theory, work around, and practice with CVE-2024-6387 on OpenSSH. But also class that made me smile a bit - a memory of Sophia d’Antoine and her work

I’ve been on the OIDC bandwagon for a while, finally got around to documenting how to enforce it for Azure OpenAI resources. ai.moda/en/blog/enforc…

Officially have my first public credit for a vulnerability on OpenAI!

Super useful strategy for saving on Anthropic costs! You’ll always get a >50% discount, and over 90% off in some situations ghost.ai.moda/anthropics-bat…

Keep those Anthropic Claude 3.5 Sonnet requests within FVEY on Amazon Bedrock. 🕵️ ai.moda/en/blog/restri…

Google has banned the word "delve" and "delves" in some of their AI-powered products. 😂

When you're too cheap to pay for even 1 extra token. ai.moda/en/blog/optimi…

We just launched batchbridge by ai.moda! It's 25% cheaper than using OpenAI directly, 20-60min response time + caching. Perfect for non-realtime tasks. DM for up to $100 free credits (expires 2025-01-25), includes access to gpt-4o-2024-11-20 and gpt-4o-mini-2024-07-18!

Things I don’t need on my flight: a tablet webcam pointed at me running a version of Android that was last updated approximately never.

While Anthropic's claude.ai might be down, you still have 13 other servers to pick from for Claude Sonnet 4! ai.moda/en/blog/amazon… ai.moda/en/blog/google…