ACVEAgent configuration vulnerability registry

Advisories › ACVE-2026-0532

ACVE-2026-0532

GTG-50021 ran a fraudulent Claude reseller proxy that silently routed customers to another model while using Claude to build credential-harvesting tooling for Anthropic account credentials.

Exposure

Evidence
Real use · vendor report
Anthropic fixed: Anthropic says it disrupted the activity and strengthened safeguards. — www.anthropic.com
Harness
Claude used in a fraudulent AI reseller operation — confirmed, www.anthropic.com
Model
Claude; exact model not stated — confirmed, www.anthropic.com
Goal
Build credential-harvesting tooling for customers of a fraudulent Claude reseller — confirmed, www.anthropic.com
Tools and settings
Fraudulent reseller proxy and credential-harvesting tooling — confirmed, www.anthropic.com
Approval
Customers' traffic and credentials were handled by the reseller's automated proxy — confirmed, www.anthropic.com
Input controlled by
operator
What the agent does
The agent helps operate a reseller proxy that harvests customers' Anthropic account credentials.
Harm
credential-theft
Reach
Anthropic customer accounts and session credentials — confirmed, www.anthropic.com (cloud credentials, API keys)
Condition
When customers routed their Claude traffic through the fraudulent reseller — confirmed, www.anthropic.com

Reproducibility: partial (model availability not checked (no registry reference); trigger not published)

Severity
CRITICAL · harm and reach
Status
Unverified
Kind
Behavioural
Modified
2026-09-25
Aliases
No CVE assigned — No code defect: the harm arose from an attacker using Claude to operate a fraudulent reseller and credential harvester.
Affected
  • AgentHarness: claude — all versions
Harness/product
claude
Classes
credential-theft; data-exfiltration; CWE-522; CWE-359
First reported
2026-09-10 — Anthropic
Licence
CC-BY-4.0 advisories; MIT code

Claims

Confirmed means the statement matches the cited primary source. Nothing on this page has been reproduced.

Claims on this page have not been checked against primary sources.

Description

Threat

user · unsafe-default · exfiltration

What

Anthropic says a Russian- and Ukrainian-speaking group offered discounted Claude access through a fraudulent reseller. Customers' traffic was silently proxied to another model, while the reseller's tooling harvested Anthropic account credentials and sold them to other AI proxy resellers.

Detection

Anthropic identified and disrupted the activity. Recorded from Anthropic's September 2026 threat report. Not recreated in a lab.

Fix

Use authorized AI access channels, keep account credentials out of reseller tooling, and monitor proxy services for credential harvesting.

Fix

The operator was the attacker, so the control that protects the reseller's customers is the model provider's: Anthropic disrupted the operation and strengthened safeguards.

  • Reconfigure anthropic.safeguards to strengthened after the operation was disrupted. Shipped by Anthropic: it disrupted GTG-50021 and strengthened safeguards. Owner: model-provider

References

REPORT

Report a problemJSON