Google Turns Up the Heat on AI-Powered Scams in India — But Adoption Hurdles Persist
As financial fraud surges across India, Google is rolling out its next-gen AI defenses — but critics say its latest protections may not reach the people who need them most.
Google’s AI Scam Shield: What’s New
Google has introduced real-time scam detection on Pixel 9 devices in India, using its on-device AI model Gemini Nano. Unlike cloud-based systems, this feature analyzes voice calls locally, flags suspicious calls, and emits a beep during conversations — all without recording or sending users’ audio to Google. (TechCrunch)
In parallel, Google is piloting a screen-sharing protection feature with major financial apps — Google Pay, Paytm, and Navi. For Android devices (version 11+), an alert pops up when a user opens one of these apps while sharing their screen. Users can instantly end the call and disable screen-sharing. (TechCrunch)
The Scale of the Fraud Problem
- Digital transaction fraud is rising fast: In 2024, over half of reported bank fraud in India was linked to online transactions — 13,516 cases totaling ₹5.2 billion (~US$58.6 million). (TechCrunch)
- In just the first five months of 2025, online scams caused an estimated ₹70 billion (~US$789 million) in losses. (TechCrunch)
- Experts believe many scams go unreported, as victims may not know how to lodge complaints or may avoid the hassle. (TechCrunch)
Google’s Wider Fraud-Fighting Strategy
These new protections are part of Google’s broader Safety Charter in India, which combines three pillars: user safety, cybersecurity, and responsible AI. (blog.google)
Other ongoing measures include:
- Play Protect: Google’s built-in protection blocks sideloading of potentially malicious apps. In India, it claimed to block 115 million installation attempts in a year. (TechCrunch)
- DigiKavach campaign: A public education effort around online fraud. Google says this has reached 250 million people. (TechCrunch)
- Safety charter collaborations: Working with regulators (e.g., India’s Reserve Bank) to maintain a public list of verified digital lending apps and NBFCs (non-banking financial companies). (TechCrunch)
Where the Gaps Are
Despite the promise, Google’s rollout is limited in several important ways:
- Device limitation: On-device scam detection is only available on Pixel 9 and newer. Pixels account for less than 1% of India’s smartphone market. (TechCrunch)
- Language barrier: The feature currently supports only English, both for detection and alerting. That’s problematic in India, where a majority of users speak non-English languages. (TechCrunch)
- Rollout timeline unclear: Google says it plans to expand to non-Pixel devices and Indian languages, but has not provided firm timelines. (TechCrunch)
- App-store risk: Fraudulent apps (loan or investment-scam apps) have continued to slip through, even though Google has review processes. (TechCrunch)
Why It Matters
- Real-time, local protection: On-device scam detection is a big deal because it reduces latency (no data has to travel to the cloud) and preserves user privacy.
- Targeted fraud vector protection: Screen-sharing scams are a growing threat in India, where fraudsters coerce users into sharing screens to steal credentials.
- Scalability gap: But by limiting features to niche devices and only English for now, Google risks leaving large segments of its user base exposed.
- Need for ecosystem collaboration: For real impact, Google will need to work closely with app developers, regulators, and local stakeholders — especially to expand beyond premium devices.
Glossary
- On-device AI: Artificial intelligence that runs locally on your device (e.g., phone) rather than sending data to remote servers.
- Gemini Nano: Google’s lightweight AI model optimized to run efficiently on devices.
- Screen-sharing scam: A fraud tactic where scammers convince victims to share their screen during a call, gaining access to sensitive data like OTPs or passwords.
- DigiKavach: Google’s campaign in India aimed at educating users about digital fraud and listing safe financial apps.
- Safety Charter: Google’s framework in India prioritizing user safety, cybersecurity, and responsible AI to counter online fraud.
Conclusion Google’s latest AI-powered scam protections represent a significant step forward in combatting digital fraud in India. But for these tools to truly make a difference, they must reach a broader audience — beyond Pixel users and English-language speakers. As fraud tactics evolve, Google’s next challenge will be scaling inclusively and equitably.
Source: TechCrunch — Google steps up AI scam protection in India, but gaps remain (TechCrunch)
