ALTV!ST

@[email protected]
1 Followers
23 Following
14 Posts
RE and bug hunting. Not affiliated with any franchise, brand, or organization. All opinions are mine.
Bloghttps://altvi.st
GitHubhttps://github.com/altvist
Mail[email protected]
ALTV!ST3d ago
Using #claude for #cybersec tasks is a total pain. Their restrictions are dumb, they trigger even if I do 100% defensive research, e.g. RE of a malicious APK.
Show threadALTV!STJun 15

They closed the report. So, the timeline is https://altvi.st/a-bug-in-apple-audiotoolbox-that-leads-to-heap-oob-read/#timeline for now.

Well, I should write a separate post and share my experience with the #apple #security #bounty program.

A bug in Apple Audio Toolbox that leads to heap OOB read | ALTV!ST

ALTV!STJun 14

A random fun fact: iMelody is a mono beep-beep-beep ringtone format introduced by Sony Ericsson in early 00s. #apple CoreAudio on the newest macOS still contains iMelody parser, and the parser is vulnerable (OOB read) 🤷‍♂️ I even don't know if I should report it to Apple.

#apple #vulnerability #ringtone #sonyericsson

ALTV!STJun 14

RE: https://infosec.exchange/@altvist/116710041220058870

It’s hard to believe, but #apple reopened the #vulnerability report again after that post (writeup + poc) in my blog. So the report is in review for the third time. What a circus 🤦‍♂️

ALTV!STJun 13

RE: https://infosec.exchange/@altvist/116722279476392134

Well, this is it https://altvi.st/a-bug-in-apple-audiotoolbox-that-leads-to-heap-oob-read/

The PoC: https://github.com/altvist/apple-audio-toolbox-oob-read-poc

#apple #vulnerability #blog

ALTV!STJun 9

I'm going to share a (non-weaponized) PoC + writeup for a couple bugs in AudioToolboxCore soon. Chained together, they lead to a heap OOB read. Not patched at the moment.

#vulnerability #apple

ALTV!STJun 7

#apple: silently closes my #vulnerability report and moves it to the status "We are unable to identify a security issue in your report"

me: thank you for your review, if it's not a vulnerability, I can publishing the writeup + PoC in my blog without worrying about ethical concerns

#apple: silently reopens the report and moves to the status "We’re reviewing your report"

ALTV!STMay 10
Recently I’ve seen a bunch of articles like “we told #Claude to find vulnerabilities and make us millionaires, and it found 100500 of them.” I got curious and tried it, but of course it doesn’t work like that. AI turned out to be effective at finding vulnerabilities only if you explain what exactly to look for and at least roughly where. Then it really does a good job.