Google said Android’s anti-fraud defenses protect users from more than 10 billion calls and messages suspected of being malicious each month, at a time when scammers are using AI tools to create more convincing schemes. The company added that continuous safety checks for the RCS service blocked more than 100 million suspicious numbers from using the service over the past month, preventing potential scams from being sent before reaching users.
This was reported in a blog post published by Lyubov Farafonova, Director of Product for Phone by Google; Alberto Pastor Nieto, Senior Product Manager for Google Messages and RCS spam and abuse prevention; and Vijay Pareek, Director of Android Messaging Trust and Safety, coinciding with the conclusion of Cybersecurity Awareness Month. The post included survey results and two reviews of mobile scam protection, in addition to a new report explaining how modern messaging scams are organized and how they are detected.
Greater Confidence According to a Google and YouGov Survey
According to a survey conducted by Google and YouGov among smartphone users in the United States, India, and Brazil, Android users said they received fewer scam messages and felt more confident in their device protection. Android users were 58% more likely to say they had not received any scam messages during the previous week, while the figure rose to 96% when Pixel users were compared with iPhone owners.
By contrast, iOS users were 65% more likely than Android users to report receiving three or more scam messages within a week. When iPhone and Pixel were compared, iPhone users were 136% more likely to report receiving a high number of scam messages. Android users were also 20% more likely to describe their device protection as “very effective” or “extremely effective,” while iPhone users were 150% more likely to say their device protection was completely ineffective at stopping phone scams.
The study’s footnote indicates a sample of 5,100 participants, with 1,700 in each country, consisting of adults who use their phones daily and had experienced an attempted scam or fraud by phone. Google stated in the summary that the number of participants was 5,000.
Evaluations of AI-Powered Protection
Google said an evaluation conducted by Counterpoint Research in October 2025, covering the Pixel 10 Pro, iPhone 17 Pro, Samsung Galaxy S25 Ultra, OnePlus 13, and Motorola Razr+ 2025, found that Android provides AI-powered protection across ten key areas, including email protection, browsing protection, and on-device behavioral protection, compared with only two areas in iOS.
In a sponsored evaluation conducted by the cybersecurity company Leviathan Security Group, covering the iPhone 17 Pro, Moto Razr+ 2025, Pixel 10 Pro, and Samsung Galaxy Z Fold 7, the analysis concluded that Android phones, led by the Pixel 10 Pro, provide the highest default level of protection against scams. The report focused on call screening, scam detection, and authentication of real-time warnings during calls.
Protection Layers in Messages and Calls
- Google Messages: It filters known spam based on the sender’s reputation and the message’s content, moving suspicious messages to the “Spam & blocked” folder. Scam Detection also uses on-device AI to analyze incoming messages from unknown senders and identify conversational scam patterns, such as the “pig butchering” scheme, while displaying real-time warnings and blocking suspicious links.
- Phone by Google: It blocks known spam calls, while Call Screen can answer on the user’s behalf to identify scam callers. If the user answers, Scam Detection continues analyzing suspicious conversational patterns and displaying warnings on the device, without saving or transmitting the call’s content.
- System protection: Android blocks high-risk actions during a call, such as installing untrusted apps or disabling security settings, and alerts users when their screen is being shared without their knowledge. These capabilities work together with real-time app scanning through Google Play Protect and enhanced Safe Browsing in Chrome using large language models.
Google explained that its evaluations are based on free features enabled by default, and that some functions may not be available in all countries. It also noted that accuracy and availability may vary, meaning that comparison results need to be read in the context of each study’s methodology and conditions.