Google is rolling out cutting-edge AI-powered defenses in India to tackle the surging threat of digital fraud. Drawing on security innovations from its Bengaluru hub, the tech giant is debuting on-device call screening to block scammers on Pixel 9 phones and piloting in-app alerts designed to disrupt real-time screen-sharing cons targeting users of financial platforms like Google Pay, Paytm, and Navi. While these measures represent a significant leap in securing mobile transactions in a country with enormous digital scale, subtle but critical vulnerabilities persist across Android and its broader app ecosystem.
How Google’s AI is Countering Scams
The centerpiece of Google’s update is real-time scam detection for Pixel 9 devices. Using the Gemini Nano AI model, this feature operates entirely on-device, scanning incoming calls from unknown numbers for patterns typical of phone scams. If suspicious activity is detected, the user receives a warning mid-conversation, highlighted by an audible beep. Importantly, audio is neither recorded nor transmitted to Google’s servers, preserving privacy. Currently, this opt-in feature is in beta for U.S. users and is slated for Indian release next.
In addition, Google has launched a pilot with major fintech players to counteract screen-sharing scams—a fast-rising fraud vector in India. Scammers often trick victims into sharing their screens to steal sensitive information like one-time passwords or account PINs. The new pilot automatically detects suspicious screen-sharing during financial app use and can alert the user or even block risky actions in real time. This build-out follows successful trials in the United Kingdom and continued collaboration with the Reserve Bank of India.
Though Google says it intends to make call-based scam protection available on non-Pixel Android smartphones, a timeline hasn’t been announced. This is a notable gap since Pixel devices account for only a small share of India’s vast device market.
Why India Remains Highly Vulnerable to Digital Fraud
India’s rapid embrace of mobile-first finance has come with new risks. In 2024, over half of all bank fraud occurred digitally, leading to losses in the billions of rupees. The Ministry of Home Affairs estimates online scams cost upwards of ₹70 billion just in the first five months of 2025. Experts believe many incidents go unreported, suggesting the numbers may be even higher.
Widespread adoption of the Unified Payments Interface and easy digital onboarding have opened formal banking to tens of millions of new users, but also created fertile ground for scammers. Tactics like voice phishing, QR code scams, fake investment offers, and abusive remote-access tools remain common, exploiting linguistic diversity and mass smartphone usage. Law enforcement agencies are ramping up responses, but fraud strategies are evolving too quickly for traditional takedown efforts to keep pace.
Google’s Progress and Early Results
Google’s security efforts extend well beyond phone calls. Play Protect, for example, now aggressively filters out loan and utility apps known to request high-risk permissions—a favorite tactic among fraudsters. So far this year, more than 115 million attempted installations have been blocked. Meanwhile, Google Pay’s risk models surface over a million transfer warnings each week for suspected fraud.
To bolster user awareness, Google’s DigiKavach digital safety campaign has reached more than 250 million people, and the company has partnered with the RBI to publish a verified list of legitimate digital lending apps. In parallel, Google’s new Enhanced Phone Number Verification seeks to replace vulnerable SMS-based OTPs with more secure, SIM-based authentication.
Where the Gaps Remain
Despite these advances, significant blind spots endure. Fraudulent or malicious apps can still slip past app store reviews, especially through clones and predatory lending services. Many users sideload APKs shared over messaging platforms, bypassing Play Store safeguards entirely. While Play Protect helps, those who disable security checks or grant broad permissions remain at risk.
The new call-screening feature currently protects only against calls from unfamiliar numbers, leaving users vulnerable to scams originating from spoofed or compromised contacts. Moreover, India’s linguistic diversity, code-mixed conversations, and constantly evolving scam tactics present ongoing challenges for algorithm accuracy. Finally, Google has yet to announce when these protections will reach the full spectrum of Android users or address threats from VoIP and in-app calls, which are becoming more common among fraudsters.
What Regulators and Banks Expect Next
Indian regulators want more robust default security for financial transactions, such as tighter screen-capture controls on banking apps, tamper-proof device authentication, and better friction or delay on abnormal transfers. Rapid takedown of malicious accounts and apps, increasingly coordinated between the RBI and cybercrime agencies, is a growing priority. Furthermore, banks urge more transparency on false positives, language support, and behavioral triggers to fine-tune fraud detection without hindering legitimate users.
What to Watch as Google Expands Anti-Scam Tools
In the coming months, the focus will be on the rollout of call-based scam detection beyond Pixel devices, broader deployment of screen-sharing protections across India’s leading financial apps, and evidence that these interventions are truly reducing fraud. Building partnerships with carrier networks and expanding language support will be vital.
For users, the essentials remain clear:
- Never share your screen during financial transactions.
- Avoid installing apps from links, even if they come from people you know.
- Check a lender’s legitimacy via the RBI’s official list.
- Report suspected fraud immediately by calling 1930.
Google’s latest AI-powered tools are a substantial step forward, but India’s battle against digital scams will require deeper, multi-layered defenses, faster responses, and continued disruption of fraudsters’ most-used channels.
