The Trigger in the Haystack: Extracting and Reconstructing LLM Backdoor Triggers

Detecting whether a model has been poisoned is a longstanding problem in AI security. In this work, we present a practical scanner for identifying sleeper agent-style backdoors in causal language models. Our approach relies on two key findings: first, sleeper agents tend to memorize poisoning dat...

Salt Typhoon APT Group: What Public Sector Leaders and Defenders Should Know

The Rapid7 Threat Focus: Salt Typhoon report profiles one of the most sophisticated and persistent state-sponsored threat actors operating today. Salt Typhoon, a Chinese espionage advanced persistent threat APT group linked to the Ministry of State Security MSS, has spent years infiltrating globa...