WhatsApp and Messenger add new warnings to help older people avoid online scams (techcrunch.com)

Meta rolled out new scam-detection safeguards in WhatsApp and Messenger aimed at protecting older users from social-engineering attacks. WhatsApp will now show an in-call warning when a user attempts to share their screen with an unknown contact — a common tactic scammers use to extract bank details or verification codes. Messenger is testing automated suspicious-message detection that flags inbound chats, displays a cautionary message (“be cautious, you could be at risk of losing money”), lists common scam patterns (e.g., work‑from‑home or easy‑cash pitches) and offers quick actions to block or report; flagged content can be routed to AI for further review. Meta says these measures come amid a broader cleanup that disrupted about 8 million scam accounts in H1 2025 and removed some 21,000 fake customer‑support pages tied to operations in places like Myanmar, Laos, Cambodia, the UAE and the Philippines. The company also joined the National Elder Fraud Coordination Center to coordinate with law enforcement and industry partners. For the AI/ML community this is a practical example of applying real‑time classification and UX intervention to reduce financial harm: models must detect subtle social‑engineering cues, run with low latency in messaging flows, and balance false positives with user privacy (especially when messages are sent to AI review). The effort highlights demand for specialized datasets and models to identify elder‑targeted scams, the operational challenge of scaling protections across languages and regions, and the tradeoffs between proactive automated defenses and users’ privacy expectations.