ChatGPT flags fake adviser; victim reports Rs 67.10 lakh fraud

AI system

System: ChatGPT
Model: Not assessed
Release: Not assessed
Application developer: OpenAI
Role: Not assessed
Configuration: Not assessed
Inputs and outputs: Not assessed
Autonomy: Not assessed
Workflow: A victim querying a large language model about a company and its representatives after withdrawals were refused. Public complaints and credential inconsistencies about the scheme that the model could surface.
Human oversight: Not assessed
Relation to AI: Both reports attribute the discovery to ChatGPT's response flagging complaints and credential inconsistencies when the victim queried the firm. That came after his own suspicion at refused withdrawals. The exact output is not reproduced. OpenAI had no role beyond providing the consumer product.
AI contribution: Both reports attribute the discovery to ChatGPT's response flagging complaints and credential inconsistencies when the victim queried the firm. That came after his own suspicion at refused withdrawals. The exact output is not reproduced. OpenAI had no role beyond providing the consumer product.

Outcome identifiers: Not assessed
Benefit: After withdrawals were refused, a 77-year-old Binnypet resident checked the firm's credentials with ChatGPT, which flagged complaints and inconsistencies; he reported the Rs 67.10 lakh fraud to Ben...
Category: Reduced Risk
Status: Not assessed
Beneficiary groups: People. Society. Organizations.
Beneficiaries: Not assessed
Reach: Not assessed
Distribution and access: Not assessed
Duration: Not assessed
Expected?: Yes
Costs and harms: Not assessed

Participants: Developer: OpenAI. Operator: The victim, using the consumer app himself.
Jurisdictions: Benefit: Bengaluru, India. Developer: United States.
Ownership and control: Not assessed
Dependencies: Not assessed
Cross-border connections: Not assessed
Geopolitical implications: Not assessed

Conditions for replication: Suspicion arising before further payments. Model output that points to specific red flags rather than generic caution. Police registration for any tracing or recovery.
Replication attempts: Not assessed
Later findings: Reports as of January 2026 say a case was registered and the money trail is being traced; no arrest or recovery was found.
Availability: Not assessed
Institutional decision: Not assessed
Grounds: Not assessed
Action and result: Not assessed
Success criterion: Not assessed

Assessed benefits: People. Society. Organizations.
Depth: People: 2. Society: Not assessed. Organizations: Not assessed.
Reasons: People: The victim recognized the fraud and filed a police complaint after ChatGPT's response. Deccan Herald says that this stopped further payments. No money was recovered in the reports read. Society: No outcome beyond the single complaint is documented. Organizations: The police registered a case. No organizational outcome is documented.
Observed or intended: Not assessed
Rubric: Not assessed
Assessor and date: Not assessed
Alternative assessments: Not assessed
Fill: Independently reported

Kind: Not assessed
Related entries: Not assessed
Field: Investment fraud. Pre-IPO share scam.
Purpose: General-purpose conversational assistant that answers users' questions.
Dates: Date basis unassessed: 2026 (precision unassessed). Updated: 21 October 2026.
Stage: Not assessed
Version: SHA-256: f6c0a087d5b2…