One guy. No team. Just Claude. He built a doxxing platform that exposed thousands of people.

A recent Anthropic threat intelligence report, published on September 10, 2026, highlights a concerning case where a single individual, utilizing AI, developed a doxxing platform. This platform successfully breached over 14 of 42 targeted organizations, compromising approximately 140,000 records, including donor lists, member data, and political affiliations, in one breach alone. This incident raises questions about whether current organizational defenses are adequate for the evolving threat landscape posed by AI-powered attackers.

This report uniquely details how one individual, using Claude, achieved breaches and data theft on a scale previously requiring a team, unlike earlier solo efforts.