Proofpoint research: AI is making many attackers worse, not better
Despite headlines suggesting artificial intelligence is upgrading cyberattacks, Proofpoint threat researcher Selena Larson tells Dark Reading she is seeing the opposite in real time, with several attacker groups getting worse at their jobs when they lean on AI. She points to TA4903, a five-year-old credential-phishing and business-email-compromise operation that once convincingly impersonated the US government with polished phishing PDFs and QR codes, and which since March has used AI at every stage of its attack chain, producing weaker results. She also cites TA2725, once a notable player in Brazilian banking trojans, which she describes as having turned into a "slop factory." She notes that the widely reported finding that 84% of ransomware attacks use some form of AI relies on self-reported survey data and doesn't prove that using it makes attacks more dangerous. For a business, the observation doesn't mean less risk, it means the quality of an attack isn't always tied to whether AI sits behind it, and defense fundamentals stay the same.
Sources
This summary was written automatically with AI from the public sources listed above and checked by an independent AI reviewer before publication. For full details, read the original source.