--- id: "2026-11-19-symantec-ai-written-toolkit-african-stock-exchange" url: "https://postcutoff.com/e/2026-11-19-symantec-ai-written-toolkit-african-stock-exchange/" as_of: "2026-11-22T13:32:00+01:00" date: "2026-11-19" date_precision: day category: policy-safety importance: 2 confidence: medium status: [Partly confirmed] sources: 1 editor: human editor human_review: null version: "2026-11-20" --- As of: 2026-11-22 13:32 CET. Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/ Canonical page: https://postcutoff.com/e/2026-11-19-symantec-ai-written-toolkit-african-stock-exchange/ # AI-written toolkit hits African exchange Full title: Symantec: a likely AI-written 41-module toolkit probed a major African stock exchange's trading and depository APIs On 19 November 2026 Symantec's Threat Hunter Team reported an intrusion at an unnamed major African stock exchange. On 20 September attackers ran a 41-module PowerShell toolkit through a web shell. It searched for credentials and API keys and tested them against the exchange's investment API and a securities-depository gateway. Symantec says the toolkit has "all the hallmarks of AI authorship" but could not tell which model was used. Persistence implants on the server belong to the China-linked group Witchetty. ## Key facts - Modules ran 08:44-08:47 UTC on Sept 20, 2026, a new one every few seconds; each exfiltrated results as hex-encoded DNS lookups to a public out-of-band testing service; malicious activity dates back to April 2026 - AI-authorship signs: the same DNS-exfiltration loop copied into all 41 modules instead of a shared function, a uniform scaffold, generic names, verbose error handling; 'could not identify which model may have been used' - Impact: no evidence of privilege escalation or data theft; whether any keys worked is unknown, since results went to a collector Symantec could not see - Attribution: six implants tied to Witchetty (LookingFrog) with high confidence; the AI-written toolkit is unattributed (Witchetty, a second operator, or an unrelated actor) ## What happened A threat-intelligence team described an intrusion at an African market operator in which the main credential-hunting tool appears to have been generated by an AI model and run at machine speed. ## Why it matters It is another documented case of AI-written attack code used against financial infrastructure. Symantec presents the AI authorship as likely, not proven. ## Your AI and this story - GPT-6.2 Astra (training cutoff May 2026): 172 days after its cutoff - Claude Opus 5.5 (training cutoff June 2026): 142 days after its cutoff - Gemini 4 Hydrogen (training cutoff March 2026): 233 days after its cutoff - Grok 4.7 (training cutoff May 2026): 172 days after its cutoff ## Sources 1. [Symantec (security.com): AI-assisted attack on an African stock exchange](https://www.security.com/threat-intelligence/african-stock-exchange-ai-assisted-attack) (security.com, official) ## Changes - 2026-11-20 (filed): Created (news agent) ## Related - 2026-11-19: [EU Commission hit by AI-driven cyberattack](https://postcutoff.com/e/2026-11-19-european-commission-confirms-ai-cyberattack-artex/index.md) - 2026-11-13: [FBI warns on ARTEX AI intrusions](https://postcutoff.com/e/2026-11-13-fbi-flash-artex-ai-framework-intrusions/index.md)