RE: https://infosec.exchange/@trailofbits/116363081880526734
Trail of Bits published a really interesting audit report on Meta's confidential computing protection for WhatsApp's AI support. One of the findings is an AML injection attack, which I wrote about in a blog post a few weeks ago: https://katexochen.aro.bz/posts/badaml/
There are many other interesting findings, and a lot to learn from them. I really appreciate that they are sharing the full report.