AI-assisted attacks are noisy, not smart: same recon command 400 times, says researcher
matthew_d_green · x · 2026-11-09
Reverse engineer PandaRE shares field observations: AI-assisted attacks aren't novel — just noisy. A human runs a recon command once; an agent will retry the same command 400 times and leaves telltale patterns like echo markers wrapping commands for output parsing.
Good news for defenders: the basics still work — volume and behavior give agents away. AI isn't making attackers smarter, just louder and slightly faster.
More from AGI Musings
- Takeoff debate: concave curves don't prove diminishing returns block superintelligence — austinc3301 · 2026-11-09
- Exponential AI progress is exponential investment with sublinear returns — world-changing, not god — dhadfieldmenell · 2026-11-09
- Takeoff debate: concave curves are by construction, not evidence of diminishing returns — austinc3301 · 2026-11-09
- Derya Tabak: People are still massively underestimating AI — abundant superintelligence changes everything — DeryaTR_ · 2026-11-09
- Counter to Sutton's "LLMs aren't bitter lesson enough": pretraining won via Hayekian aggregation — akbirthko · 2026-11-09
- fkasummer: 2028 hinges on a new paradigm for intelligence — transformers, backprop all up for grabs — fkasummer · 2026-11-09