Android AI Scam Protection to Fight Online Fraud Google is expanding Android’s AI-powered scam protection with new security features designed to help users identify fraudulent messages, suspicious links,
Google is expanding Android’s AI-powered scam protection with new security features designed to help users identify fraudulent messages, suspicious links, fake calls, and attempts to steal sensitive information. The latest improvements aim to make smartphones safer as cybercriminals increasingly use artificial intelligence and sophisticated social engineering techniques to deceive people.
Announced on October 14, 2026, the new protections include an enhanced security mode in Google Messages, smarter link checks, safeguards against call-forwarding fraud, and automatic protection for sensitive information during certain screen-sharing sessions. Google is also working with Android device manufacturers to extend scam alerts to more users. (Google Security Blog)
Mobile scams have become more sophisticated than traditional spam messages. Criminals may impersonate banks, delivery companies, government agencies, employers, or family members to convince people to share personal information or transfer money.
Some attackers use fake websites, manipulated caller IDs, and AI-generated voices to make fraudulent requests appear genuine. Others build trust through long conversations before introducing fake investment opportunities or urgent financial emergencies.
These tactics can be difficult to identify using basic spam filters alone. Android’s expanded protections use AI-based analysis and additional security checks to help identify suspicious behaviour earlier.
Google says its existing scam detection features in Phone by Google and Google Messages flag more than five million suspected scam calls and messages each month. (Google Security Blog)
One of the main additions is an optional Enhanced protection mode in Google Messages. It is designed to help identify sophisticated scams that may develop gradually through conversations with unknown senders.
With the enhanced setting enabled, scam detection can use Google’s Gemini cloud models through Private AI Compute to analyse complex message context. This supplements existing on-device detection and can help identify suspicious intentions that might not be obvious from a single message.
For example, a conversation may begin with an apparently harmless job offer, investment discussion, or friendly introduction before the sender asks for money or sensitive information.
Google says Private AI Compute is designed to process this analysis in a secure environment that protects user privacy.
The feature is being rolled out in stages, so availability may vary by device and region.
Fraudulent messages often contain links to websites that imitate legitimate banking, shopping, or account-login pages. These websites may attempt to steal passwords, payment information, or one-time passwords.
Existing protections can identify known malicious links, but scammers frequently create new websites that have not yet been catalogued.
With Enhanced protection enabled, Gemini Nano can help assess unfamiliar links on supported devices using real-time web signals. Android can then provide additional warnings to help users pause before opening a potentially dangerous website.
These warnings are useful, but users should still verify the website address and avoid entering sensitive information after following an unexpected link.
Google is also working with Android manufacturers to extend scam alerts in chat notifications beyond supported Pixel devices.
These alerts can warn users about potentially suspicious messages directly from the notification area, helping them identify possible scams before opening a messaging application.
Google says it plans to bring this protection to hundreds of millions of additional users through partnerships with manufacturers including Motorola, OnePlus, OPPO, realme, TECNO, vivo, Xiaomi, and others.
The rollout is staged, and not every Android smartphone will receive the same capabilities at the same time.
Call forwarding can be useful when people need calls redirected to another number. However, scammers may try to manipulate victims into enabling forwarding so that incoming calls, including account-verification calls, reach a number controlled by the attacker.
Android is introducing safeguards against applications secretly enabling call forwarding in the background. It can also warn users when they are tricked into entering certain call-forwarding codes manually.
These measures are intended to make it harder for criminals to redirect calls without the user’s informed consent.
Users should never enter unfamiliar codes or change phone settings simply because an unexpected caller claims it is necessary to secure a bank account.
Screen sharing is commonly used for remote assistance, meetings, and technical support. Unfortunately, scammers may exploit it to watch users enter passwords, banking information, or payment details.
Android is extending existing safeguards so that sensitive login fields and payment screens in Chrome and Android WebView can be automatically hidden during supported live screen-sharing sessions.
This provides an additional layer of protection when sensitive information appears on the screen.
However, it does not make every screen-sharing session safe. Users should be cautious when someone they do not know asks them to share their screen, install a remote-access application, or open a banking app during a phone call.
Android combines several layers of security rather than relying on one AI feature.
| Security feature | How it helps |
|---|---|
| AI scam detection | Identifies suspicious patterns in calls or conversations |
| Enhanced Google Messages protection | Analyses complex message context for possible scams |
| AI-assisted link checks | Helps assess unfamiliar links for potential danger |
| Notification scam alerts | Warns users about suspicious messages on supported devices |
| Call-forwarding safeguards | Helps prevent unauthorised call redirection |
| Screen-sharing privacy | Hides certain sensitive fields during supported sessions |
Some protections work directly on the device, while Enhanced protection in Google Messages can use secure cloud processing. Features and privacy details differ, so users should review the information provided for their particular device and setting.
AI-powered protection can help detect suspicious activity, but users still play an important role in preventing fraud.
Follow these practical precautions:
Not immediately. Google is rolling out the new features in stages, and availability depends on the device model, Android version, supported applications, language, and country.
Some AI capabilities require compatible hardware, while others depend on software updates or regional availability. Users should check official Android and device-manufacturer announcements for confirmation.
The expanded rollout is intended to extend protections to more than just Pixel users, but it does not mean that every feature will appear on every Android device.
Android’s expanded AI-powered scam protection reflects the growing need for stronger mobile security as fraudsters adopt increasingly convincing techniques. By improving message analysis, checking suspicious links, protecting screen-sharing sessions, and addressing call-forwarding abuse, Google aims to help users recognise threats before they cause harm.
These features can provide valuable assistance, but no automated system can identify every scam. Keeping devices updated, verifying unexpected requests independently, and refusing to share passwords or OTPs remain essential security practices.
As AI continues to change both cybersecurity and online fraud, combining intelligent detection with informed user behaviour will be important for protecting personal information and financial accounts.
It is a collection of security features that use AI and other detection methods to help identify suspicious calls, fraudulent messages, dangerous links, and certain scam-related activities.
It is an optional security setting designed to improve detection of complex scams by analysing suspicious conversations and using additional link-safety checks on supported devices.
No. AI can identify many suspicious patterns, but scammers constantly change their techniques. Users should verify unexpected financial requests and avoid sharing sensitive information.
No. Availability varies by device, software version, application, and region. Some features are being rolled out gradually.
Keep software updated, avoid suspicious links, never share OTPs or PINs, verify callers independently, and report suspected cybercrime promptly.
