Product design question

How do you solve the fake messages problem in WhatsApp/other messaging platforms?

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Product sense on trust and safety in an encrypted messaging product, where message content cannot be read.

Clarify "fake messages" means misinformation forwards, not spam, since WhatsApp is end-to-end encrypted and content is private.
Identify the mechanism of harm: mass-forwarded messages spreading unverified claims, especially in family and community groups.
Prioritize forward-chain behavior over content, since that is visible metadata WhatsApp can act on without breaking encryption.
Propose interventions: the existing five-chat forward limit, a "forwarded many times" label, and a tap-through friction step for highly forwarded messages.
Weigh false positives: legitimate viral content like memes or event info gets caught by the same friction.
Define success as reduction in highly-forwarded message volume, since content-based fake-message counts can't be measured under encryption.

Correctly identifies that encryption blocks content-based detection and pivots to metadata, forward count and chain depth, as the real lever.
References the actual shipped mitigation, the five-chat forward limit, as a baseline, then proposes incremental friction beyond it.
Names the real tradeoff: over-restricting forwards hurts legitimate viral sharing and user experience.
Proposes measuring highly-forwarded message volume as a proxy metric since content can't be inspected.

Proposing content moderation or scanning, which breaks end-to-end encryption and is not feasible.
Ignoring the false-positive cost to legitimate sharing.
Not acknowledging what WhatsApp has already shipped.

Chapters of the AI PM course that teach what this question tests.