Microsoft unveils AI security tools it says outperform competing platforms
Microsoft says tools cost less than competing ones and outperform them, too. https://arstechnica.com/security/2026/07/microsoft-unveils-ai-security-tools-it-says-outperform-competing-platforms/?utm_brand=arstechnica&utm_social-type=owned&utm_source=mastodon&utm_medium=social
Comments (3)
@arstechnica But they also consume water and energy resources more disproportionately than any previous technology mankind has ever devised.
They're burning the planet to death in service to making a few more million dollars before they die too. It's a death cult and they're taking us all with them.
@arstechnica but does it really? Too early to say for sure https://infosec.exchange/@BenAveling/116957390389825149
Ben Aveling (@BenAveling@infosec.exchange)
A record breaking patch Tuesday from Microsoft. 622 vulnerabilities, 72 of which are marked critical, and Chromium has another 422 vulnerabilities, which also affect Microsoft Edge. There are two possibilities: AI is really good at finding vulnerabilities in old code. AI is really good at putting vulnerabilities into new code. What happens next? If, given a bit of time, the number of bugs that are being found goes down, or if it goes up further but then goes back down to normal, then that will suggest that these are old bugs that have been flushed out by AI. But. If the number of bugs being found doesn’t go back down to something like what used to be normal, then that would suggest these are new bugs that are being added by AI. And either way, at least for companies using vibe-coding, old or new does not matter, the lesson for either is the same: unless we are a lot better than Microsoft we have a lot of vulnerabilities and we had better find them before someone else does.
infosec.exchange@arstechnica That... would not be hard to do, considering that all the other AI security tools suck donkey dongs. Saying that as someone whose employer recently installed a competing tool and now we're hastily triaging security vulnerabilities that the AI is hallucinating.