Google, Singapore team up on AI tools to catch scams before they land

Singapore's cyber defenders are getting early access to Google's newest security arsenal. The two sides are working together under Fairwind Program, an initiative the tech giant rolled out a month ago, with the tie-up made public at Singapore International Cyber Week.

The arrangement builds on earlier joint work. Google and the city-state previously teamed up to add anti-scam safeguards to Android handsets, and the government's GovTech agency is now co-developing an on-device scam detector with Google DeepMind for the same platform.

Under Fairwind, defenders gain tools such as the Gemini 4 Argon model and the CodeMender harness. Google says these let teams locate, confirm and repair weaknesses far quicker than the weeks such work traditionally takes, and patching ahead of time should keep essential services running and shield them from attack.

The on-device detector is the more consumer-facing piece. Instead of shipping message content to remote servers, the analysis happens on the handset, which means even encrypted conversations can be checked, something cloud scanning cannot do today. Using local data for training, it should catch impersonation of banks, public agencies and well-known local figures, and keep learning as fraud tactics shift.

For the industry, the direction of travel is clear: smaller models running locally, tuned to one market's language and institutions, rather than one-size-fits-all cloud filters. The privacy angle matters too, since users get warnings in real time without surrendering their chats. Google's own numbers hint at the scale of the problem, with more than 3.3 million attempts to install suspect apps blocked across over 770,000 devices through its app store's enhanced fraud protection, alongside 68,000 unique apps flagged for possible financial misuse.

Why it matters for Singapore: The island is a frequent target for fraudsters, which makes it a useful proving ground for privacy-first AI. The partnership fits Singapore's ambition to be a trusted AI hub with clear rules of the road, and it hands local firms, researchers and regulators a live example of agentic security tools that could shape both procurement and policy. It also strengthens the case for growing homegrown AI talent and tooling.