Google steps up AI scam protection in India, but gaps remain

In a proactive move to combat digital fraud in India, Google is leveraging artificial intelligence to bolster its defenses. The tech giant has introduced on-device scam detection for its Pixel 9 smartphones and has launched new alerts for financial applications that involve screen sharing. As more individuals in India embrace digital transactions, the incidence of online fraud has surged. According to the Reserve Bank of India (RBI), digital fraud cases accounted for over half of all reported banking fraud in 2024, totaling 13,516 incidents and leading to losses exceeding ₹5.2 billion (approximately $58.61 million). The Ministry of Home Affairs further reported that online scams resulted in estimated losses of ₹70 billion (around $789 million) in just the first five months of 2025. Many victims, however, may not report these scams due to uncertainty about the complaint process or a desire to avoid further scrutiny. On Thursday, Google unveiled its enhanced real-time scam detection feature, which utilizes Gemini Nano technology to analyze calls directly on the device. This feature identifies potential fraud without recording conversations or sending data to Google’s servers. Initially available only on calls from unknown numbers, the feature is turned off by default and notifies users of potential fraud with a beep during calls. It first launched in the U.S. as a beta for English-speaking Pixel 9 users back in March. Currently, Google's on-device scam detection is restricted to Pixel 9 and later models in India, and it only serves English-speaking users. This limitation poses a challenge in a country where a significant portion of smartphone users primarily speak non-English languages, as Android holds nearly 96% of the smartphone market according to Statcounter, yet Pixel devices represented less than 1% of that share in 2024. Google has acknowledged this language barrier and is working on expanding detection capabilities to non-Pixel Android devices, although a timeline has yet to be established. Additionally, Google has initiated a pilot program with financial applications such as Navi, Paytm, and Google Pay to address screen-sharing scams. In these schemes, fraudsters often coax victims into sharing their screens to gain access to sensitive information such as one-time passwords and PINs during calls. The feature, which was first highlighted at Google I/O in May, has been tested in the U.K. and will be available to users with devices running Android 11 or newer, providing alerts in Indian languages along with a simple option to terminate calls and halt screen sharing. Furthermore, Google’s Play Protect service has been active in combating predatory loan applications in India by preventing the sideloading of apps that request sensitive permissions commonly exploited for fraudulent purposes. This year alone, Play Protect has blocked over 115 million installation attempts. Google Pay also plays a crucial role in this effort, issuing more than a million warnings weekly for transactions flagged as potentially fraudulent. The company has launched the DigiKavach campaign to raise awareness about digital fraud, reaching over 250 million individuals. In collaboration with the Reserve Bank of India, Google has published a list of authorized digital lending apps to reduce the presence of malicious actors in the ecosystem. Despite these initiatives, significant gaps remain in Google's ability to combat digital fraud effectively. Similar to Apple, Google has faced scrutiny for allowing misleading and fraudulent applications to remain on its app store, despite existing review processes intended to filter out such submissions. Security researchers and law enforcement have previously flagged investment and loan applications utilized in scams that continued to be available on the Play Store without timely intervention. These ongoing challenges highlight the complexities Google faces in managing a market where it holds a dominant position.