The Government Technology Agency of Singapore (GovTech) and Google DeepMind have announced a partnership to develop on-device AI tools for detecting conversational scams. The project utilizes lightweight versions of Google’s Gemma models to analyze private messages in real time.
A key feature of the system is its local processing design. The analysis occurs entirely on the user's device, ensuring that no data is transmitted to the cloud. This approach is intended to address privacy concerns while providing early intervention before financial losses occur.
The initiative was introduced by Minister for Digital Development and Information Josephine Teo at Singapore International Cyber Week. The new tool will complement existing government defenses, including the SATIS system, which blocks approximately 20 000 malicious websites monthly, and the rMSE system, which scans over 600 000 websites daily.
Specific details regarding the launch timeline and performance metrics, such as detection accuracy, remain undisclosed. Future implementation will depend on collaboration with messaging platforms and device manufacturers to integrate the technology into consumer applications.
