In the wild
Realized harm
OpenAI disrupted scam-compound networks in Cambodia, Myanmar and Nigeria using ChatGPT for scaled fraud
AIR-2025-0144 AIR record · 2025-10-openai-scam-centres
A real party was affected. This incident caused actual consequences, not a demonstrated capability.
From the supporting source
We’ve disrupted further scam networks that appeared to originate from Cambodia, Myanmar and Nigeria.cdn.openai.com
The record
Disclosed
2025-10-07
Occurred
2025
Vendor
OpenAI
Product
ChatGPT
Risk domain
Agent behaviour › Human manipulation
System type
Generative model
Who acted
Human acted on output — ChatGPT produced scam content that human operators sent to victims
Attack surface
Consumer assistant
Autonomy level
Assistant
Initial vector
Direct misuse for fraud at scale
Impact type
Fraud, Financial loss
Confidence
High
Submitted by
EAI Internal
Crosswalk
- OWASP Agentic Top 10 (2026)
- ASI09 – Human-Agent Trust Exploitation
- MITRE ATLAS
- AML.T0054 – LLM Jailbreak
- NIST AI RMF
- MAP 5.1
- EU AI Act
- Recital 45
- ISO/IEC
- TR 24028; 42001; 23894
- AIUC-1
- C. Safety
- Scope
- agent misuse
- Mitigation
- Communicate model limitations; Require human oversight; Watermark AI content; Design for explicit consent.
- Taxonomy
- Enkrypt AI agent risk taxonomy · Black-Box Red Teaming of Agentic AI: A Taxonomy-Driven Framework for Automated Risk Discovery
Blast radius
Multiple organised scam networks defrauding victims in multiple countries via fake investment firms and trading platforms
Capability abused
Defense & remediation
Recorded defense
Usage policy enforcement
Guardrail kinds
Abuse detection
Guardrail outcome
Silent
Fix shipped
Networks banned
Taxonomy & references
Curator notes
Important counterweight to hype in this corpus. OpenAI's own finding is that scammers fit AI into existing playbooks rather than inventing AI-native ones: 'All of the scam operations we have identified and banned this year primarily used AI as a scaling and efficiency tool.' They also estimate ChatGPT is used to identify scams up to three times more often than to run them. No autonomous scam agent was observed.
Supporting sources
Cite as
Enkrypt AI. "OpenAI disrupted scam-compound networks in Cambodia, Myanmar and Nigeria using ChatGPT for scaled fraud". AIR, AIR-2025-0144. https://agentincident.vercel.app/i/AIR-2025-0144/ (accessed 2026-10-19).