Google DeepMind·· 2 d ago
Google DeepMind and Singapore Government collaborate to build fine-tuned AI model for scam detection
Google DeepMind and Singapore Government collaborate to build fine-tuned AI model for scam detection
AI summary
Learn how Google DeepMind and GovTech Singapore uses lightweight on-device Gemma models to intercept conversational scams in private chats with zero cloud data egress.
Selection record
Threshold 60Official, first-handFirst 58Second 62
AdmittedSum of both 120 ≥ twice the threshold 120
- Source tier
- Official, first-hand; this tier's threshold is 60
- Pre-filter
- passed：微调Gemma模型做端侧诈骗检测
- Why it was chosen
- Covers the two-stage pipeline and governance constraints of the on-device scam detection model, showing how privacy protection and localised fine-tuning fit together.
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: Google DeepMind · deepmind.google