serverinspector retweeted
Unfortunately, I can't use any of @AnthropicAI latest models like Opus 5.5 or Sonnet 5.5 for anything defense related. Both models refused to work in a pure defense posture.
Right now @OpenAI is the only AI company helping defenders, and I appreciate it.
serverinspector retweeted
what omp got wrong, what omp² changes. why "a while loop around a fetch" is systems software in a trenchcoat: stencil.so/blog/harness-play…
more people should use models to find cves and help secure things, it also brings immense joy, but if you, someone with experience is still stuck on how many CVEs u found rather than gargantuan impact ones i suggest u retire and do SEO optimization or something
serverinspector retweeted
A record-breaking year for Microsoft's Bounty Programs!
This year, Microsoft awarded more than $20 million to 562 security researchers, the highest total payout and largest number of researchers recognized in program history.
Read the full blog to learn more about the impact of the global security research community: aka.ms/microsoft-bounty-year…
🤖 Made with AI
serverinspector retweeted
So proud of our security team! They caught, contained & publicly disclosed an attack unlike anything we've seen before, and did it at record speed.
Also massively grateful to @Zai_org: they shared GLM5.2 as open weights (for free!) with the world and it became a key part of our defense.
This is day one for cybersecurity in the age of agents & we're all learning that secrecy is not the answer & that all defenders (not just a few selected ones) everywhere need more powerful models without restrictions, especially open ones!
Very weird that we're at a point where you can just point LLMs at repos and tell it to go wild. I feel like things will only change more wildly in security research from now on. Insane finding.
Seems that wp2shell PoCs are now floating around the internet, so we've published our blog post including our research methodology for finding the bug as well as a deep dive into the chain itself - slcyber.io/research-center/e…
Insane hacker/feature type matchup here. 4x super effective. Great bugs, great tricks. Research Review. youtu.be/SEgUNWHXxrM
We've published a new blog post by RyotaK @ryotkak
He discovered 8 methods to bypass safety mechanisms in Claude Code, leading to arbitrary command execution.
We recommend updating to v1.0.93 or later to fix this vulnerability (CVE-2025-66032).
flatt.tech/research/posts/pw…
Replying to @serverinspector @reach_vb
Actually never mind, I believe I know where I got this from after some people reached out. Not sure if I can disclose so 🤐.
serverinspector retweeted
Posting some appreciation for @msftsecresponse right here. I see the amount of work they put in to keeping users safe and researchers happy. That does not mean things will always work out the way you want. It means they do their best for each of the (many many many) reports.
Agreed. MSRC is easily one of the best teams I've interacted with. Communication is never an issue in my cases, and I hope that continues.
Replying to @HaifeiLi
As I said previously, the MSRC and all security folks I engaged with are mostly very nice in person, the security improvements in Microsoft software and services are what we could see in our labs and during our daily research, the ~17M yearly bounty payout are real, and many more. IMO MSRC has been an absolutely leader and has basically defined what the vendor Security Responses look like today (I recall a lot of *SRCs). There’s definitely zero reason for Microsoft to kill all the decades-long good efforts and community relationships in one single post (can’t imagine that😅).
There were and there will be sometimes very hard to deal with for some cases, no doubt. If things go bad, I will complain bad. But with more effective and direct communications, I think (at least I hope) we can improve continuously.
Overall, I’m personally very happy to see this clarification coming out and hopefully this drama can be resolved peacefully.
serverinspector retweeted
Over the past several days, we have been listening to the conversation around coordinated disclosure and the relationship between security researchers and vendors. We recognize that this relationship is both critical and, at times, fragile. We deeply value the security community, and will continue to take your feedback seriously.
To be clear about our approach to legal matters, we have no intention to pursue action against individuals conducting or publishing their security research. When an individual breaks the law and engages in malicious activity causing real harm to our customers, we will work with law enforcement as appropriate.
We recognize the work that goes into researching and submitting a vulnerability. We are committed to approaching every interaction with transparency, clear communication, and professionalism. We continue to believe strongly in Coordinated Vulnerability Disclosure as the foundation for protecting customers and improving our products. Each year we process a high volume of vulnerability reports. That volume continues to grow and will continue with the rise of AI-enabled research. We acknowledge that some interactions have fallen short and are working to learn from them.
Many of us have experience on both sides of this work, as researchers reporting vulnerabilities and as responders triaging and assessing them. That perspective informs how we approach this feedback and the importance we place on getting it right, particularly as the volume and complexity of research continues to grow.
The security community plays a vital role in helping us protect customers. We are committed to maintaining a constructive and respectful relationship and growing together. We know that, given the nature of this work, there will at times be misunderstandings. We remain committed to engaging in good faith and to providing a respectful and professional experience for all researchers, regardless of past interactions.
Readers added context they thought people might want to know
Contrary to this claim, Microsoft previously threatened legal action via its Digital Crimes Unit against researcher Nightmare Eclipse for publishing unpatched vulnerabilities. pcmag.com/news/microsoft…
An issue with the current harnesses available is the fact that they are "coding harnesses".
I feel like this is a misstep in how agents are supposed to evolve, it started on the wrong path and everyone just kept walking on it.
A great harness should do anything, not just code.
serverinspector retweeted
Every security researcher starts somewhere. For Felix, it began with arcades and led all the way to Azure.
Now a Microsoft MVR and Zero Day Quest qualifier, Felix shares the persistence and problem-solving mindset that shaped his path into security research in our latest blog: msft.it/6014Qn7B4
serverinspector retweeted
“You don’t pick the bugs. The bugs pick you.”
Meet Wouter (@wtm_offensi), Microsoft MVR and Zero Day Quest 2026 qualifier, and read his security research journey: msft.it/6016QHeaC
serverinspector retweeted
I'm looking for a new role in web application penetration testing. I've spent the last three years in offensive security, with bug bounties on the side and speaking at conferences. Last year, I spoke at DEF CON on enterprise integration servers and spoke at GRR CON and BSides. I've got talks coming up at CactusCon, BSides Seattle, and CypherCon on postMessage attacks. Remote preferred. If you're hiring or know someone who is, shoot me a message.
serverinspector retweeted
Close friend of mine is looking to find a new role. 4 years experience in SOC/IRT and he has a wide variety of experience in this area. He’s based in Ireland. I can personally vouch for him. If anyone has a role in mind for him I’d be very thankful! 😊 Please RT.
serverinspector retweeted
I wrote a post that has a bunch of me and my friends arguing lolllll
do you think Prompt Injection is a bug?
josephthacker.com/ai/2025/11…


























