Google on Thursday announced that it was unfolding with a new artificial intelligence (AI) that works in combating Chrome, Search and Android scams.
The technical giant said he would start using Gemini nanoits a great linguistic model on your device (LLM) to improve Safe view In Chrome 137 on the working stacks.
“The approach to the device gives an instant understanding of risky sites and allows us to offer protection, even from scams that have not been seen before. LLM Gemini Nano is ideal for this use due to the ability to translate a variety – Note.
Google noted that it is already using this approach to AI to fight distance technical support that often seeks to deceive users to develop their personal or financial information under the pretext of a non -existent problem with the computer.
This works by evaluating the web page using LLM for potential signals that are a symbol of scammers, such as using API Blocking Keyboard. The security signals are then extracted and transmitted to a safe view to determine whether the scam is.
“In addition to ensuring that the llm is only triggered sparingly and run locally on the device, we careful manage resource consumption by conscription the number tokens used, running the process asynchronous Browser Activity, and Implementing Throttling and Quota Enforcement Mechanisms to Limit Gpu Usage, “Jasika Bawa, Andy Lim, and Xinghui Lu of the Google Chrome Security Team – Note.
Google said he intends to expand this feature to detect other types of fraudsters, including related packages and unpaid fences. This feature is expected to be transferred to Chrome on Android this year.
As part of the Google announcement, it also showed that it expanded its scam detection systems that operate on AI to capture 20 times more deceptive pages and block such pages from search results, reducing the schemes that represent the airlines more than 80%, and those that mimic official resources, such as visas and public services.
Finally, Google stated that it launches a new Warning Function on Android, which uses a machine learning model to warn users about unwanted notifications sent to malicious sites that are aimed at fooling them in downloading suspicious software or providing sensitive data.
“This new feature uses machine training on the device to discover and warn you about potentially deceptive or spamic notifications, giving you an additional level control over your device,” Hanna Boana “Chrome Sachery and Sarah Krakowak Wing – Note.
“When Chrome denotes the notice, you will see the name of the site that sends the notice, a message that warns that the notice content is potentially deceptive either spam, and the opportunity to either cancel the subscription from the site or see the content of the designation.”
Features come a little over two months after Google Rolled out Features of detection of AI scam in app messages for Android. Last year’s company disclosed Similar ways caused the scam.
Updates also come when Google seems to be preparing Extended Defense Function In Android 16 that in a sense, Apple’s mirrors Disconnecting JavaScript, turning off 2G connections and activating a number of default security features such as blocking thefts, offline lock, safe Android viewing, spam protection in messages.
Google was also noticed by working on a function to detect scams report Earlier this week.
