探索我们的下一代 AI 系统
Gemini
Gemini 构建智能体
Gemini Omni 从任何事物创造任何事物
Nano Banana 创建和编辑精细图像
Gemini Audio 交谈、创建和控制音频
专用模型
Veo 生成带音频的电影级视频
Genie 3 生成并探索互动世界
Lyria 生成高保真音乐和音频
Gemini Robotics 感知、推理、使用工具并互动
开放模型
我们最新的 AI 突破以及来自实验室的更新
突破
SIMA 2 一个在虚拟 3D 世界中与你一起游玩、推理和学习的智能体
Genie 3 生成并探索互动世界
AlphaGo 掌握围棋游戏
Gemini Robotics 感知、推理、使用工具并互动
了解更多
用 AI 开启发现的新时代
突破
AlphaFold 高精度预测蛋白质结构
AlphaGenome 使用 AI 理解人类基因组
WeatherNext 快速且准确的 AI 天气预报
AlphaEarth 以前所未有的细节绘制我们的星球
AlphaEvolve 为数学和计算应用设计高级算法
了解更多
我们的使命是负责任地构建 AI,以造福人类
责任 通过主动安全确保 AI 安全,甚至应对不断演变的威胁新闻 了解我们最新的 AI 突破、项目和更新职业 我们正在寻找希望对世界产生真正积极影响的人
了解更多
教育我们的国家 AI 合作伙伴关系加速器项目播客DeepMind Institute
模型
探索我们的下一代 AI 系统
Gemini
Gemini 构建智能体
Gemini Omni 从任何事物创造任何事物
Nano Banana 创建和编辑精细图像
Gemini Audio 交谈、创建和控制音频
专用模型
Veo 生成带音频的电影级视频
Genie 3 生成并探索交互式世界
Lyria 生成高保真音乐和音频
Gemini Robotics 感知、推理、使用工具并交互
开放模型
研究
我们最新的 AI 突破和实验室动态
突破
SIMA 2 一个与你一起游玩、推理和学习的智能体
Genie 3 生成并探索交互式世界
AlphaGo 掌握围棋游戏
Gemini Robotics 感知、推理、使用工具并交互
了解更多
科学
用 AI 开启发现的新时代
突破
AlphaFold 高精度预测蛋白质结构
AlphaGenome 使用 AI 理解人类基因组
WeatherNext 快速准确的 AI 天气预报
AlphaEarth 以前所未有的细节绘制我们的星球
AlphaEvolve 为数学和计算应用设计高级算法
了解更多
关于
我们的使命是负责任地构建 AI,造福人类
了解更多
教育我们的 AI 国家合作伙伴关系加速器项目播客DeepMind 研究所
责任 通过主动安全确保 AI 安全,甚至应对不断演变的威胁新闻 探索我们最新的 AI 突破、项目和动态职业 我们正在寻找希望对世界产生真正积极影响的人
Google AI 了解我们的所有 AIGoogle DeepMind 探索 AI 的前沿Google Labs 体验我们的 AI 实验Google Research 探索我们的研究
产品与应用
Gemini 应用 与 Gemini 对话Google AI Studio 使用我们的下一代模型进行构建Google Antigravity 我们的智能体开发平台
2026年10月14日 责任与安全
Google DeepMind 与新加坡政府合作构建用于诈骗检测的微调 AI 模型
JJ Geewax 和 Miguel de Andrés
- [x]
分享
与新加坡政府科技局(GovTech Singapore)合作,在不损害隐私的前提下保护用户。
诈骗,例如投资欺诈、虚假工作机会和冒充公职人员,占新加坡报告犯罪的一半以上,并自2019年以来造成超过40亿新元的损失。与此同时,诈骗手法已从简单的网络钓鱼转向复杂的社会工程学,攻击者通过令人信服的对话操纵受害者,然后索要钱财。此类诈骗现在占所有损失的80%。
为解决这一问题,Google DeepMind 和新加坡政府科技局(GovTech Singapore)作为我们与新加坡政府国家 AI 伙伴关系的一部分,一直在合作。我们共同微调了一个保护隐私的设备端诈骗检测模型。
不损害隐私的设备端 AI 保护
通用模型往往难以应对本地欺骗手法,使诈骗检测更加困难。为解决这一问题,Google 的轻量级 Gemma 模型基于本地诈骗对话进行了微调。这些对话涵盖冒充公共机构、银行和本地人物,以及流行的区域骗局。这确保微调后的模型在部署到公民安装的应用程序中时,能够区分良性社交闲聊和真正的操纵行为。
与依赖云端服务器端扫描消息的传统威胁检测不同,该模型可以直接在设备上处理对话数据。当集成到应用程序中时,这将允许更快地分析对话模式以检测操纵意图,并帮助公民在发生财务或凭证泄露之前采取行动。
该模型在应用程序中的推出将谨慎分阶段进行,并与新加坡政府合作开发。部署后,该模型只能在用户同意的情况下访问设备上的对话,并遵守该应用程序的使用条款。由于模型完全在设备上运行,新加坡政府和 Google 都无法通过其部署访问用户的对话。
打击诈骗不应意味着牺牲个人隐私。与 Google DeepMind 合作,我们将先进的 AI 保护直接带到公民的设备上。这使我们能够实时捕捉高度针对性的本地诈骗企图,同时保持个人通信的私密性。
Mark Chen
新加坡政府科技局信任与安全总监
底层揭秘:本地语境与边缘效率
当模型检测到操纵性对话模式时,它会直接在用户设备上触发实时警告,建议他们在分享敏感信息或资金之前暂停并保持谨慎。
在现代智能手机上运行复杂的诈骗分类器模型并非易事,电池效率是一个关键关注领域。为此,我们依赖一个两阶段流水线:
- 阶段 1:一个轻量、高效的分流器对入站消息进行筛查,功耗极低,专注于捕捉潜在威胁。
- 阶段 2:使用我们微调后的 Gemma 模型进行上下文分析,我们会分析更大范围的对话快照,以确认威胁或排除误报。
建立在严格治理之上
这项合作在新加坡国际网络周(SICW 2026)上展示,为消息平台在用户保护与端到端加密之间取得平衡提供了一个范例。整个开发过程始终保持严格治理:
- 主权沙盒:微调在新加坡政府科技局管理和维护的 GCP 环境中进行,使训练数据保持隔离。
- 不使用公民数据:训练严格排除私人或可识别个人身份的信息,仅使用源自诈骗报告和情报源的匿名数据集。
- 压力测试:该模型针对合成对话和本地语料库进行了测试,以覆盖多种诈骗类别和不断演变的冒充手法。
设备端安全的未来
该项目展示了 AI 如何在不损害安全或隐私的情况下帮助打击犯罪。它还表明政府与模型提供方可以如何合作,打造服务于公共利益的科技。借助这一轻量级设备端模型,我们证明了本地实时诈骗检测的能力,同时确保聊天交互保持安全。我们期待很快分享更多关于我们与新加坡政府合作的信息,并探索类似技术在其他国家部署的可能性。
相关文章
加强新加坡的 AI 未来:一项新的国家合作伙伴关系
2026 年 5 月 责任与安全
以 AI 加速规划释放英国住房建设潜力
2026 年 6 月 责任与安全
宣布我们与大韩民国的合作伙伴关系
2026 年 4 月 责任与安全
通过 AI 驱动的科学与教育加速印度发现进程
2026 年 2 月 责任与安全
加强我们与英国政府的合作伙伴关系,以支持 AI 时代的繁荣与安全
2025 年 12 月 责任与安全
深化我们与英国 AI 安全研究所的合作伙伴关系
2025 年 12 月 责任与安全
Google DeepMind 支持美国能源部推进 Genesis:一项加速创新与科学发现的国家使命
2025 年 12 月 科学
关注我们
注册以获取我们最新创新的动态
我接受 Google 的条款和条件,并确认我的信息将按照 Google 隐私政策 使用。
注册
负责任地构建 AI,造福人类
模型
GeminiGemini OmniNano BananaGemini Audio GemmaGenieLyriaVeo
研究
科学
AlphaFoldAlphaGenomeWeatherNextAlphaEarthAlphaEvolve
产品
Gemini 应用Google AI StudioGoogle Antigravity
了解更多
关于新闻职业机会人工智能国家合作伙伴关系加速器项目播客DeepMind 研究所
Cookie 管理控制
Explore our next generation AI systems
Gemini
Gemini Build intelligent agents
Gemini Omni Create anything from anything
Nano Banana Create and edit detailed images
Gemini Audio Talk, create and control audio
Specialized models
Veo Generate cinematic video with audio
Genie 3 Generate and explore interactive worlds
Lyria Generate high fidelity music and audio
Gemini Robotics Perceive, reason, use tools and interact
Open models
Gemma Build responsible AI applications at scale
Our latest AI breakthroughs and updates from the lab
Breakthroughs
SIMA 2 An agent that plays, reasons, and learns with you
Genie 3 Generate and explore interactive worlds
AlphaGo Mastering the game of Go
Gemini Robotics Perceive, reason, use tools and interact
Learn more
EvalsPublicationsResponsibilityFrontier safety
Unlocking a new era of discovery with AI
Breakthroughs
AlphaFold Predict protein structures with high accuracy
AlphaGenome Using AI to understand the human genome
WeatherNext Fast and accurate AI weather forecasting
AlphaEarth Map our planet in unprecedented detail
AlphaEvolve Design advanced algorithms for math and applications in computing
Learn more
Gemini for ScienceExperimental ToolsScience Skills
Our mission is to build AI responsibly to benefit humanity
Responsibility Ensuring AI safety through proactive security, even against evolving threatsNews Discover our latest AI breakthroughs, projects, and updatesCareers We’re looking for people who want to make a real, positive impact on the world
Learn more
EducationOur National Partnerships for AIAccelerator programsThe PodcastDeepMind Institute
Models
Explore our next generation AI systems
Gemini
Gemini Build intelligent agents
Gemini Omni Create anything from anything
Nano Banana Create and edit detailed images
Gemini Audio Talk, create and control audio
Specialized models
Veo Generate cinematic video with audio
Genie 3 Generate and explore interactive worlds
Lyria Generate high fidelity music and audio
Gemini Robotics Perceive, reason, use tools and interact
Open models
Gemma Build responsible AI applications at scale
Research
Our latest AI breakthroughs and updates from the lab
Breakthroughs
SIMA 2 An agent that plays, reasons, and learns with you
Genie 3 Generate and explore interactive worlds
AlphaGo Mastering the game of Go
Gemini Robotics Perceive, reason, use tools and interact
Learn more
EvalsPublicationsResponsibilityFrontier safety
Science
Unlocking a new era of discovery with AI
Breakthroughs
AlphaFold Predict protein structures with high accuracy
AlphaGenome Using AI to understand the human genome
WeatherNext Fast and accurate AI weather forecasting
AlphaEarth Map our planet in unprecedented detail
AlphaEvolve Design advanced algorithms for math and applications in computing
Learn more
Gemini for ScienceExperimental ToolsScience Skills
About
Our mission is to build AI responsibly to benefit humanity
Learn more
EducationOur National Partnerships for AIAccelerator programsThe PodcastDeepMind Institute
Responsibility Ensuring AI safety through proactive security, even against evolving threatsNews Discover our latest AI breakthroughs, projects, and updatesCareers We’re looking for people who want to make a real, positive impact on the world
Google AI Learn about all our AIGoogle DeepMind Explore the frontier of AIGoogle Labs Try our AI experimentsGoogle Research Explore our research
Products and apps
Gemini app Chat with GeminiGoogle AI Studio Build with our next-gen modelsGoogle Antigravity Our agentic development platform
October 14, 2026 Responsibility & Safety
Google DeepMind and Singapore Government collaborate to build fine-tuned AI model for scam detection
JJ Geewax and Miguel de Andrés
- [x]
Share
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.
Fighting scams shouldn't mean sacrificing personal privacy. Working with Google DeepMind, we're bringing advanced AI protection directly onto citizens' devices. This allows us to catch highly targeted, local scam attempts in real time while keeping personal communications private.
Mark Chen
Director, Trust and Safety, GovTech Singapore
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.
Related posts
Strengthening Singapore’s AI Future: A New National Partnership
May 2026 Responsibility & Safety
Unlocking UK house-building with AI-accelerated planning
June 2026 Responsibility & Safety
Announcing our partnership with the Republic of Korea
April 2026 Responsibility & Safety
Accelerating discovery in India through AI-powered science and education
February 2026 Responsibility & Safety
Strengthening our partnership with the UK government to support prosperity and security in the AI era
December 2025 Responsibility & Safety
Deepening our partnership with the UK AI Security Institute
December 2025 Responsibility & Safety
Google DeepMind supports U.S. Department of Energy on Genesis: a national mission to accelerate innovation and scientific discovery
December 2025 Science
Follow us
Sign up for updates on our latest innovations
I accept Google's Terms and Conditions and acknowledge that my information will be used in accordance with Google's Privacy Policy.
Sign up
Build AI responsibly to benefit humanity
Models
GeminiGemini OmniNano BananaGemini Audio GemmaGenieLyriaVeo
Research
Gemini RoboticsBreakthroughsEvalsPublicationsFrontier safetyResponsibility
Science
AlphaFoldAlphaGenomeWeatherNextAlphaEarthAlphaEvolve
Products
Gemini appGoogle AI StudioGoogle Antigravity
Learn more
AboutNewsCareersNational Partnerships for AIAccelerator programsThe PodcastDeepMind Institute
Cookies management controls
本文内容采集自官方网站,排版和翻译可能与原页面存在差异。
阅读官方全文