Anthropic-Disclosed AI-Orchestrated Campaign (Sept 2025) — Status: resolved

NoteThe claim states the campaign's status is 'resolved (as of 2025-09)' with '4 successful breaches confirmed out of ~30 targets.' The source confirms: (1) the campaign occurred in mid-September 2025, (2) roughly 30 targets were targeted, (3) it succeeded 'in a small number of cases' (which aligns with 4 breaches), (4) Anthropic terminated access (GTG-1002 is not mentioned by name, but the source states 'we banned accounts as they were identified'), and (5) it is 'the first documented case' of AI-orchestrated cyberattack. However, the source does not explicitly state the campaign status as 'resolved' as of September 2025—it was detected in mid-September but the investigation continued 'over the following ten days,' with the public disclosure occurring in November 2025. The exact number '4 successful breaches' is not stated; the source only says 'a small number of cases' and 'succeeded in a small number of cases,' which is consistent with but not precisely confirmed as 4. The claim's temporal framing ('as of 2025-09') is ambiguous—if it means the campaign occurred in September, that's confirmed; if it means the status was 'resolved' by end of September, that's unverifiable from the source.