[ctx]ctxseccontext security: attacks on the harness and context layer of AI agents, plus exploited-now news, hunting and playbooks.

← Industry News · markdown

official (U.S. Attorney SDNY press release / indictment, Nov 5); AP summary of the charging papers

SDNY charges three men with using AI voice and video deepfakes of CEOs and attorneys to steal about $129 million from companies worldwide

  • How it worked (per charging papers / AP): email claiming to be from the CEO asking for help with a confidential transaction; then a phone or video call where the defendants used AI to mimic the CEO's voice and sometimes image; then a handoff to an "attorney" at a major firm who was also an impersonator directing large wires abroad.
  • Defendants: Soall (nicknamed "Rocky Balboa" in reporting) mainly played UK attorneys and kept a handwritten "Lawyer Call" script; Hazan ("Superman Cash") has a prior 2008 NY bank-fraud conspiracy conviction; Sheetrit ("Fat Boy"). They allegedly sat in Spain, the UK, Israel and the UAE.
  • Why it matters for security teams: this is classic BEC with a deepfake upgrade. Out-of-band verification on any CEO/GC wire request, dual control for high-value transfers, and skepticism of unexpected video calls that demand secrecy still beat tooling alone.

Sources: U.S. Attorney SDNY: Three foreign nationals arrested and charged with scheme to use super intelligence to defraud companies of $129 million · SDNY on X (Nov 5) · SecurityWeek / AP: 3 Charged in Scheme to Use AI to Impersonate CEOs and Attorneys to Steal $129 Million