WhatsApp has begun a limited beta rollout of Scam Alert, an optional feature that uses an on-device machine learning model to flag suspicious messages from non-contacts.
The company says the tool is designed to work alongside end-to-end encryption rather than around it, with classification happening entirely on the device and no automatic reporting to WhatsApp or parent company Meta.
Once a user enables the feature, a model is downloaded to the device and evaluates incoming messages for patterns associated with known scams, drawing on conversational structure and linguistic cues.
If a message is flagged, only the recipient sees a warning inside the chat, but the sender is not notified. The user can block the contact, report the message, ignore the warning, or mark the conversation as trusted so future alerts are suppressed. Users who mark a chat as trusted can separately choose to share the last five messages received to help refine the model’s accuracy.
To prevent a scenario where a specific model could be pushed to a specific person, WhatsApp says every model release must be logged on a third-party, append-only transparency ledger before it can be distributed.
Each release is accompanied by a manifest of SHA-256 hashes covering the model weights and related files, with the manifest digest signed using Ed25519 keys held by Cloudflare rather than Meta. Devices verify this signature, cross-check it against the ledger, and confirm the downloaded files match the published hashes before the model is allowed to run.
Because no message content leaves the device, WhatsApp still needs a way to gauge whether the feature is functioning correctly. For that, it built what it calls a confidential federated analytics pipeline, which collects only two categories of data: counts of how often warnings were triggered and counts of what action users took afterward, such as blocking or marking a chat as trusted.
WhatsApp outlines a threat model covering external attackers, compromised infrastructure insiders, and supply-chain risk, and says the pipeline’s defenses are meant to ensure that targeting a single user’s data would require compromising the entire system.
On the verification side, users will be able to review a transparency log in the app, accessible through Account > Request Info > Scam Alert Activity, showing which messages were scanned, the outcome, and which model version made the call.
WhatsApp also says it is expanding its bug bounty program to cover Scam Alert, and researchers stress-tested the system ahead of the beta rollout.
The company characterizes the release as an early technical preview rather than a finished product, noting the feature will continue to evolve during the beta period based on researcher and user feedback before any wider rollout.
Signal announces automatic key verification
Signal has introduced automatic key verification, a new feature meant to complement its existing safety number system by confirming that no unauthorized party sits between two users in an end-to-end encrypted conversation.
Unlike manual safety number checks, which require an in-person meeting or a secondary communication channel, the new system performs verification independently through checks carried out by the user, their contact, and third-party auditors.
The feature is designed to catch scenarios where a Signal account’s public key gets swapped without the owner’s knowledge. This can happen, for instance, if an attacker compromised Signal’s infrastructure and linked a different key to a target’s phone number.
Users can trigger a check from a connection’s profile by opening ‘View Safety Number’ and tapping ‘Verify automatically’ under the Automatic Key Verification heading. A green checkmark reading ‘Encryption verified’ confirms a match.
Underpinning the feature is a system Signal calls key transparency, which records every registration, phone number change, and username change in a cryptographically verifiable log.
Cloudflare and Trail of Bits act as independent auditors of this log. All identifiers and keys in the log are obscured using a verifiable random function and a keyed hash function, so the auditors themselves never see plaintext user data.
The feature is opt-out. Users who don’t want to rely on Signal or its auditors can disable it under Privacy > Advanced > Automatic Key Verification and continue relying solely on manual safety number checks.
*added more information on WhatsApp Scam Alert testing via bug bounty program
Related: WhatsApp Rolling Out Username Feature to Bolster Phone Number Privacy
Related: WhatsApp Discloses File Spoofing, Arbitrary URL Scheme Vulnerabilities
Related: Germany Suspects Russia Is Behind Signal Phishing That Targeted Top Officials
