MalwareTech: LLM-hallucinated php2shell exploits flooding GitHub don't actually work
paul_cal · x · 2026-10-23
Security researcher MalwareTech warns that LLMs are hallucinating large volumes of non-functional php2shell exploits, with clueless users celebrating "successful" hacks. After consulting top exploit developers, he concludes that anyone claiming AI "easily" produced a working exploit is either a world-class expert with extreme imposter syndrome — or falling for confident-but-wrong LLM output. His honeypot network is being bombarded with these fake exploits, which aren't remotely close to how real exploitation works. Paul Calfighiu adds that human experts remain valuable for at least a few more months.
More from Fun
- BREAD: the AI world's new backronym for retrieve-embed-analyse-done — intellectronica · 2026-10-23
- Dev Jokes: You Can Only Relax Using Claude on Stuff That Doesn't Matter — louisvarge · 2026-10-23
- How would AI paint humanity? This 1689 Flemish painting is the answer — ZeroStateReflex · 2026-10-23
- Abliterated models trend again, and Nous Research is being mocked for quitting the game — realsohamparekh · 2026-10-23
- X Comedy Classic: Human-Capital Debater Endlessly Trolled With 'Do You Have a Girlfriend Yet?' — teortaxesTex · 2026-10-23
- Opus 5.5 codes an entire sci-fi movie title sequence—every frame and synthesized note — prasenx · 2026-10-23