pierce@slopsec:~$ ./identify
slopsec
sketchy shit, non-deterministic prices
AI security research and consulting for systems that hallucinate, delegate, exfiltrate, improvise, or otherwise make incident response more interesting.
hello@slopsec.org$ cat /research/focus.txt
- agentic system red teaming
- LLM application exploit research
- AI supply-chain and model-adjacent risk
- evals that fail loud enough to be useful
- security advisory for weird, fast-moving AI products
slop is the research surface: generated code, generated reasoning, generated actions, generated confidence. We poke at the places where AI systems become infrastructure before they become predictable.