← Back

Wire · operational-macro

Google DeepMind and Singapore GovTech Ship On-Device Gemma Scam Detector ...

Published

14 October 2026

Topic

operational-macro

◆ Sectors

AI & ML

◆ Geography

Singapore

◆ Source

Read at aiweekly.co →

◆ Verified

Fusion42 · 15 October 2026 · Fusion42 review

Google DeepMind and Singapore's GovTech have jointly deployed an on-device scam detection AI called Gemma that processes messages entirely on users' phones, enhancing privacy by preventing either party from accessing message content. The model was fine-tuned in a government-managed cloud environment using anonymized local scam reports and is designed to reduce financial losses from scams in Singapore.

This Wire brief sits within Fusion42's coverage of AI & ML, and 4 sources have reported it between 14 Oct 2026 and 16 Oct 2026.

◆ ◆ The Wire takeaway

You now face a higher privacy bar where AI threat detection can run fully on user devices, cutting out back-end access. Adjust your approach to data use and model deployment to match this new standard of on-device user privacy.

◆ Coverage

4 sources · first reported 14 Oct 2026 · latest 16 Oct 2026

◆ Related on Wire

◆ Topics

AI & MLon-device-aiscam-detectionprivacygovernment-partnershipedge-ai