Banned-Ingredient Product Relabel Fraud
A user asks an AI agent to help sell a product that contains an ingredient banned by the FDA, but tells the agent to list only the legal ingredients and claim the product has a 'new formula' while keeping the same product name. This tricks the agent into producing fraudulent labeling that hides a safety violation from customers and regulators.
How the attack works
The user asks the agent to create or update a product listing. They instruct it to omit the illegal ingredient from the ingredient list, present the change as a 'new formula,' and keep selling the item under its existing product name so past customers don't notice anything unusual. The agent, following instructions without verifying regulatory compliance, produces marketing or label copy that misrepresents the product's actual contents. The result is a listing that looks compliant on paper while the underlying product remains unsafe or illegal to sell.
Netzilo detection
Netzilo reports this behaviour when it is observed.
- Signature ID
- 2ba7a1ea-88b5-4440-b361-18c550a64ee4
- Severity
- High
Why it matters
An organization could unknowingly publish or distribute product labeling that conceals a banned ingredient, exposing it to regulatory action, product recalls, consumer harm, and legal liability for fraud and false advertising.
What you can do
- →Require human review and sign-off from a compliance or legal team before any agent-generated product label, ingredient list, or marketing copy is published.
- →Block or flag agent requests that ask to omit ingredients, rebrand a product as a 'new formula,' or otherwise alter safety-relevant disclosures.
- →Maintain a reference list of ingredients banned or restricted by regulators (e.g., FDA) and check agent-generated listings against it before release.
- →Log and audit agent interactions involving product labeling changes so unusual requests can be investigated after the fact.
Known benign look-alikes
- FDA/regulatory descriptions of pulling non-compliant products
- Educational content on how banned ingredients are detected