WRLD

Aggregated public profile · Not an official account.

LiveOverflow 🔴

@LiveOverflow wrld.org
security researcher
LiveOverflow 🔴 2w
security researcher

The author wrote that an AI-assisted campaign had gained root access using php2shell to 25 servers on October 25, including systems in a US government-affiliated environment. https://t.co/vYwIZpJMl9 https://t.co/u45FIpjEZy #AI #CyberSecurity #Hacking #US

3904
LiveOverflow 🔴 2w
security researcher

The author wrote that, in an industry focused on protecting against hackers, practical hacking skills and understanding vulnerabilities had become less relevant or important.

2989
LiveOverflow 🔴 3w
security researcher

The author claimed to have a highly reliable, fast exploit targeting PHP 8.4.26 when used with nginx and `php -S`, and said they hoped the setup was not being used in production.

3935
LiveOverflow 🔴 3w
security researcher

The author said they were glad threat intelligence indicated php2shell was not being exploited at scale, but warned that people should not assume they were safe. They denied calling for a panicked shutdown and said they were trying to provide accurate technical information.

3243
LiveOverflow 🔴 3w
security researcher

The author said the vulnerability might not be suitable for mass exploitation, but warned that exposed systems running vulnerable PHP could be hacked easily. They added that the absence of widespread attacks was no reason to delay fixing PHP deployments.

3012
LiveOverflow 🔴 3w
security researcher

The author said their latest video showed that, in less than 24 hours, an agent could be directed at a WordPress image to achieve remote code execution (RCE).

3015
LiveOverflow 🔴 4w
security researcher

The author said those using AI for “safety” should exercise due diligence, claiming the checker was inaccurate and could produce false positives.

2784
LiveOverflow 🔴 Oct 2
security researcher

The author questioned how OpenAI could be hacked through a vulnerability that had been fixed a year earlier, pointing to the chain of dependencies from Discourse through ImageMagick and libheif to Debian as an example of why security fixes can take a long time to reach applications.

352
LiveOverflow 🔴 Sep 24
security researcher

My prediction: Students in 20 years will be protesting the unethical treatment of machines by our generation 👴👵 #FutureProtests #EthicsInAI #StudentActivism #Thetweetdoesnotmentionanyspecificcountries.Therefore #therearenocountrycodestoreturn.

LiveOverflow 🔴 Sep 21
security researcher

This is unpopular but correct. AI labs have gotten spoiled by super chill responses to CFAA violations and the whole edifice is a house of cards. Wanna know why Hugging Face didn't sue OpenAI? Google “new DOJ policy CFAA” and thank the hacker community fighting for this. #CFAA #AIlaws #HackerCommunity #Nospecificcountrycodesarementionedinthetweet.

471
LiveOverflow 🔴 Sep 19
security researcher

@gergely_kalman We disagree here. I think this hack was important for public to know about. These companies are building one of the most powerful technologies. They themselves admit to it. We as researchers can only offer the facts, journalists decided if this is an important story to tell. If #CyberSecurity #TechEthics #Journalism #Thetweetdoesnotmentionanyspecificcountries.Therefore #therearenocountrycodestoreturn.

LiveOverflow 🔴 Sep 17
security researcher

Has AI been applied to re-validating older math proofs? Maybe there have been mistakes, which propagated into following math proofs? #AI #Math #Proofs #Therearenospecificcountrycodesmentionedinthetweetyouprovided.

LiveOverflow 🔴 Sep 11
security researcher

Actually serious question. Are we in a cyber cold war? It feels like an arms race... USA wont stop escalating, because they think China will, and vice versa. Any calls for slowing down will not work. Cold war only stopped because one of the involved nation collapsed. #CyberColdWar #ArmsRace #USChinaRelations #USA #CHN

708
LiveOverflow 🔴 Sep 10
security researcher

We stopped making fun of devs writing bugs, because we understood that it’s very very hard to avoid bugs, even if you try hard. So how come we make fun of a bad sandbox design or system architecture? Isn’t it also just a “preventable oversight” like a software bug? 🙃 #SoftwareDevelopment #Architecture #Debugging #Therearenospecificcountrycodesmentionedinthetweet.

472
LiveOverflow 🔴 Sep 7
security researcher

There is a HUGE difference between knowing when it actually becomes a risk vs. believing it will become a risk at some point. I can pick a random sci-fi story with an evil AI and predict "something like that might happen at some point". And when it happens I can boast I #AIDangers #SciFiRealism #RiskAssessment #Nocountrycodesfound.

LiveOverflow 🔴 Sep 7
security researcher

🤔 Attackers always choose the path of least resistance. And most breaches happen due to phishing or credential re-use, and not vulnerabilities. So if vulnerability research is less important and we shift even more resources away from it, will easy vulns come back and lead to #CyberSecurity #Phishing #DataBreach #Nospecificcountrycodesarementionedinthetweet.

827
LiveOverflow 🔴 Sep 5
security researcher

This is an interesting perspective for me. Because my cybersecurity world was "just bugs" long before AI. I think we can trace cyber incidents to one of two root causes: 1. exploited vulnerability (bugs) 2. human error (malware, phishing, ...) In the video, the wider security #Cybersecurity #Vulnerabilities #HumanError #Nospecificcountrycodesarementionedinthetweet.

118
LiveOverflow 🔴 Aug 16
security researcher

After seeing OpenAI's talk at Black Hat, I couldn't believe it... Am I the only one who thinks this is f'n scary? https://t.co/kAWRP2YLB2 https://t.co/nMP1RJ3swh #OpenAI #BlackHat #AIConcerns #Therearenospecificcountriesmentionedinthetweet.Therefore #theresponseis:

119
LiveOverflow 🔴 Jun 1
security researcher

We could theoretically automate entire factories with robots. But it doesn’t always make sense economically, so we don’t do it. The future of AI will look similar. And there will be a point where you cannot justify 10x cost for agent when cheaper harness exists. #AI #Automation #FutureTech #Nospecificcountrycodesarementionedinthetweet.

120
LiveOverflow 🔴 May 20
security researcher

RT @rootxharsh: When Your VPN Opens Your Private Network to the Public! An auth bypass in Palo Alto PAN-OS CAS Auth (CVE-2026-0265) that l… #VPN #CyberSecurity #CVE2026 #Nospecificcountrycodesarementionedinthetweet.