Google announced Gemini-powered fraud detection, link analysis and privacy protection tools for Android devices.
Google announced Gemini artificial intelligence-powered fraud detection, on-device connection analysis and hiding sensitive information in screen sharing features for Android users.
AI-generated summary
Google is introducing new AI-based protection layers to increase user security in the Android ecosystem.
Google announced new and improved anti-fraud and privacy protection features for Android users. Among the new tools is an optional fraud detection system powered by Gemini artificial intelligence that can be used in the Google Messages application.
According to Engadget, the Scam Detection feature to be added to the Google Messages application, when activated by the user, will evaluate the intent of the message sender by using Google's advanced Gemini models. The system will warn the user when it determines that the messages may be coming from a malicious person.
In the example shared by Google, a suspected fraudster pretends to be trying to send a photo of himself and his dog to the recipient. When the recipient indicates that the message was sent to the wrong person, the sender continues the communication and asks for the recipient's name. The new system aims to detect suspicious behavior in such conversations and warn users against possible fraud attempts.
MESSAGES WILL BE REVIEWED IN A SECURE ENVIRONMENT
Google stated that it will also implement measures to protect privacy in the analysis of messages. According to the company's statement, Android will analyze messages in a secure environment called Private AI Compute. It was stated that this environment, where messages are encrypted, was designed so that it cannot be accessed by anyone else, including Google.
In addition, Gemini Nano will analyze links in Google Messages directly on the device before the user clicks them, evaluating whether they are safe or not.
The connection security feature will be available on select devices, including the Google Pixel 10 and later models, as well as the Samsung Galaxy S26 series, Galaxy Z Flip, and Galaxy Z Fold 8.
Google has also designed new scam alerts that will give users more information and more time to decide before tapping on links from numbers that are not in its contact list. The new optional security setting is planned to be available in the coming weeks.
SCAM ALERTS ARE COMING TO MORE DEVICES
Google is expanding the scope of the Fraud Detection feature it previously offered through message notifications on Pixel devices. This feature will show alerts about possible fraud attempts directly in the message notification. This will allow users to obtain information about the risk without having to open the suspicious message.
The company announced that it is collaborating with various manufacturers, including Honor, Motorola, OPPO, OnePlus, realme, vivo and Xiaomi, to make the feature available on more devices. These manufacturers are expected to make the feature available on their devices later this year.
SENSITIVE INFORMATION WILL BE HIDDEN IN SCREEN SHARING
Another security feature announced by Google focuses on protecting sensitive information during screen sharing.
The feature, which is planned to be available to all devices using Android 15 and newer versions later this month, will hide passwords, banking information and security numbers of credit cards by default when the screen is shared with Chrome or Android WebView open.
Google is also working on a new protection system to prevent misuse of its call forwarding feature. The feature, which is planned to be available on Pixel and select other devices in early December, will prevent apps from secretly enabling call forwarding in the background.
The system will also alert users if fraudsters attempt to persuade users to enter an MMI forwarding code, a short string of numbers used to perform a specific action, to enable call forwarding on their device.
MORE THAN 3 MILLION FAKE CALLS WERE BLOCKED IN BRAZIL
Google stated that it is also cooperating with banks in Brazil against fraud attempts carried out through bank impersonation.
In this context, Android uses an authentication system that checks whether incoming calls are actually made from a bank. If the calling party fails the verification, the system ends the call and informs the user.
Google announced that more than 3 million fake searches have been blocked to date thanks to this feature, which was introduced in June. It is not yet clear when the security tool will be available in the USA. The company stated that it plans to work with more partner banks in the future.
AI outlook — possibilities, not facts
The new security tools will be available in the coming weeks and later in the year.
Very likely · Within weeks
Türkiye Artificial Intelligence Initiative (TRAI) updated its initiative map as of October 2026. With the addition of 43 new startups, the total number reached 504, exceeding the 500 threshold for the first time. Significant growth has been observed in the fields of Agentic AI, LegalTech and HR Tech.

BİM announced that it will offer LG's 65 and 55 inch QNED Mini LED television models and a 2.1 channel soundbar system in its current catalog on October 18. The prices of the products vary between 4,990 TL and 45,900 TL.

The National University of Singapore, DayOne and Cortical Labs have launched the 'Biological Data Centre' project working with neurons derived from human stem cells. 20 CL1 biocomputer units provide energy efficiency by integrating with traditional silicon chips.

NVIDIA introduced its new RTX Spark artificial intelligence and graphics architecture at a global launch held in Istanbul. More than 1,200 participants experienced new hardware and won special awards at the event at Vatan Bilgisayar Topkapı store.

It was determined that the Israeli-based company Discover-Tech was using ad technologies to infect users with spyware that does not require a click. According to the Citizen Lab report, the devices of Al Jazeera employees were targeted with this method.

Apple's new models iPhone 18 Pro and Pro Max broke a record by exceeding the total sales threshold of 4 million in the Chinese market. According to Digital Chat Station data, only the Pro Max model approached 900 thousand weekly activations.