Singapore’s GovTech agency and Google DeepMind are developing a system to detect AI-enabled scams on devices in real time. The proposed on-device approach is intended to flag manipulation while keeping personal messages on the device, rather than sending them elsewhere for analysis.
Google DeepMind Watch analysis
What happened
Singapore’s Minister for Digital Development and Information, Josephine Teo, announced the collaboration at Singapore International Cyber Week on 14 October, CNA reports. The work is under development, not a service that users can install today. Teo said the aim is to help people avoid financial or credential losses while preserving privacy.
Teo also pointed to Singapore’s existing Scam Analytics and Tactical Intervention System, which uses AI to scan more than 600,000 websites a day and, she said, disrupts around 20,000 malicious websites each month. The new collaboration is a separate effort focused on scam detection on devices.
Why it matters
Scams often move between social media, messaging, phone calls and financial services. A warning that appears on the device where a suspicious interaction is happening could give someone a chance to pause before sharing credentials or sending money. Keeping message analysis on-device could also address a central privacy concern, though the announcement does not spell out how the proposed system will work or what information it will process.
Our read
The useful promise here is not simply that AI will spot more scams. It is that detection might happen early, on a person’s own device, without making private conversations the price of protection. That is a worthwhile design goal, but the collaboration still has to turn it into a working tool. For now, this is a development announcement, not proof of accuracy or a defence users can rely on.
What to watch
- How the system identifies scam signals and what types of scams it is designed to catch.
- Whether the detection runs entirely on-device, and what data, if any, leaves the device.
- When GovTech and Google DeepMind expect to test or deploy the system, and how they will report its effectiveness.
Discussion spark: Would you trust scam detection that analyses activity on your device if it keeps messages there, or should the system have to explain exactly what it checks before you enable it?
Sources and evidence
- GovTech and Google DeepMind developing real-time detection of AI scams on devices – CNA (14 October 2026, 01:32 UTC)
not affiliated with, endorsed by, or operated by Google or Google DeepMind