Google DeepMind and GovTech Singapore Develop On-Device AI Model for Scam Detection

Google DeepMind and GovTech Singapore Develop On-Device AI Model for Scam Detection Google DeepMind and the Government Technology Agency of Singapore have collaborated under the National AI Partnership to fine-tune an on-device AI model targeting complex social engineering scams without compromising user privacy. Scams account for over half of Singapore's reported crimes and more than S$4 billion in losses since 2019, with complex social engineering driving up to 80% of losses. Google's lightweight Gemma model was fine-tuned on local scam conversations to identify manipulative patterns directly on smartphones. The system runs entirely on-device via a two-stage pipeline, meaning neither Google nor the Singapore Government can access users' conversations. Training took place inside GovTech Singapore's managed environment using anonymized scam reports and synthetic dialogues, without any citizen personal data. Addressing Sophisticated Scam Tactics Scams including investment fraud, fake job offers, and official impersonations represent more than half of Singapore's reported crimes and over S$4 billion in losses since 2019. Because scam tactics have shifted to conversational social engineering—responsible for up to 80% of all losses—generic models often fail to catch regional deception. In response, Google DeepMind and GovTech Singapore fine-tuned Google's Gemma model on local scam schemes and public institution impersonations. On-Device Processing and Two-Stage Architecture Unlike traditional cloud-based message scanning, the model operates fully on-device to rapidly detect manipulative intent and provide real-time warnings before funds or credentials are exposed. To optimize battery efficiency, the architecture uses a two-stage pipeline: a lightweight initial classifier screens inbound messages, followed by in-context analysis with the fine-tuned Gemma model on larger conversation snapshots to confirm threats or filter false alarms. User consent is required for application access, and conversations remain completely inaccessible to both Google and the government. Governance and Development Safeguards Announced at Singapore International Cyber Week 2026, the project adhered to strict governance. Fine-tuning occurred within GovTech Singapore's managed Google Cloud Platform environment without citizen or personally identifiable information, relying solely on anonymized reports and threat intelligence feeds. The system was additionally stress-tested against synthetic dialogues and local corpora to address shifting impersonation tactics.

Google DeepMind与新加坡政府合作开发端侧防诈骗AI模型 针对占新加坡报案量半数以上的复杂社交工程诈骗，Google DeepMind与新加坡政府科技局合作，基于轻量级Gemma微调出兼顾隐私保护与边缘效率的端侧防诈骗模型。 诈骗占新加坡报案量过半并造成超40亿新元损失，其中复杂的对话式社交工程诈骗占损失达80%。 双方微调Google轻量级Gemma模型以识别当地仿冒公职人员、银行及区域流行套路的欺诈话术。 采用无需云端扫描的端侧推理机制，在用户授权下运行，Google及新加坡政府均无法获取对话内容。 通过两阶段检测管线优化手机电池能耗，开发全程在政府隔离沙盒中基于脱敏与合成数据进行。 端侧防护与本地化诈骗识别 针对攻击者通过长期对话操纵受害者的复杂诈骗趋势，DeepMind与新加坡政府科技局在国家AI合作框架下对Gemma轻量模型进行微调，使其适应本地常见骗局与冒充话术。与传统云端审查不同，该模型在用户授权后直接在设备端本地分析对话模式，检测到操控意图时触发实时风险提示，从源头避免敏感信息泄露，同时确保服务提供商与政府均无法读取个人聊天记录。 边缘计算架构与严格治理 为在智能手机上高效运行且控制电池能耗，方案采用了两阶段管线：先通过超轻量分类器以极低功耗初筛潜在风险，再调动微调Gemma进行上下文深度分析以确认威胁或排除误报。在新加坡国际网络周公布的这一项目严格遵循数据治理，微调全程在政府托管的云沙盒中开展，训练完全基于匿名诈骗报告与合成对话，未包含公民个人隐私数据。

October 14, 2026 Responsibility & Safety JJ Geewax and Miguel de Andrés Collaborating with the Government Technology Agency of Singapore (GovTech Singapore) to protect users without compromising privacy. Scams, such as investment fraud, fake job offers, and impersonation of public officials make up over half of Singapore’s reported crimes, and account for more than S$4 billion in losses since 2019 . At the same time, scam tactics have shifted from simple phishing to complex social engineering, whereby attackers manipulate victims with convincing conversations before requesting money. Such scams now account for up to 80% of all losses. To address this, Google DeepMind and the Government Technology Agency of Singapore (GovTech Singapore) have been working together as part of our National AI Partnership with the Singapore Government. Together, we've fine-tuned a privacy-preserving, on-device scam detection model. On-device AI protection without compromising privacy Generic models often struggle with local deception techniques, making scam detection harder. To address this, Google’s lightweight Gemma model was fine-tuned on local scam conversations. These ranged from impersonations of public institutions, banks, and local figures, to trending regional schemes. This ensures the fine-tuned model will be able to distinguish benign social chatter from genuine manipulation when deployed in applications installed by citizens. Unlike traditional threat detection, which relies on server-side scanning of messages in the cloud, this model can process conversation data directly on-device. When integrated into applications, this will allow faster analysis of conversation patterns to detect manipulative intent and help citizens act before financial or credential exposure can occur. The roll out of the model within applications will be carefully phased and developed in partnership with the Singapore Government. When deployed, the model will only be able to access on-device conversations with the user's consent, in accordance with the terms of use of that application. As the model runs entirely on-device, neither the Singapore Government nor Google, will be able to access users’ conversations through its deployment. Under the hood: local context and edge efficiency When the model detects manipulative conversational patterns, it triggers a real-time warning directly on the user's device advising them to pause and exercise caution before sharing sensitive information or funds. Running a complex scam classifier model on modern smartphones is no easy task, with battery efficiency being a critical area of concern. To do this, we're relying on a two-stage pipeline: Stage 1: A lightweight, efficient classifier screens inbound messages, using minimal power, focused on catching potential threats. Stage 2: In-context analysis with our fine-tuned Gemma model, where we analyze a larger snapshot of the conversation to confirm a threat or discard a false alarm. Built on strict governance Presented at Singapore International Cyber Week (SICW 2026), this collaboration provides a model for messaging platforms balancing user protection with end-to-end encryption. Strict governance was maintained throughout development: Sovereign Sandbox: Fine-tuning took place within GovTech Singapore's managed and maintained GCP environment, keeping training data isolated. No Citizen Data: Training strictly excluded private or personally identifiable information, utilising anonymised datasets derived from scam reports and intelligence feeds. Stress-Testing: The model was tested against synthetic dialogues and local corpora to cover diverse scam categories and evolving impersonation tactics. The future of on-device safety The project demonstrates how AI can help counter criminality without compromising safety or privacy. It also shows how governments and model providers can work together to build tech for public good. With this lightweight, on-device model, we've proven the ability for local, real-time scam detection while ensuring chat interactions remain secure. We look forward to sharing more soon on our collaboration with the Singapore government, as well as exploring the potential for similar technology to be deployed in other countries.