--- title: "Phishing: Gmail Gets AI Integration for Protection" description: "Google has added unforeseen external reply warnings to Gmail Google agrees that about 50% - 70% of mails are spam on Gmail New updates seem to b" author: "Anonymous User" published: 2017-06-15 updated: 2018-03-20 canonical: https://www.mindstick.com/blog/11458/phishing-gmail-gets-ai-integration-for-protection category: "phishing" tags: ["google", "cybersecurity"] reading_time: 3 minutes --- # Phishing: Gmail Gets AI Integration for Protection ![Phishing: Gmail Gets AI Integration for Protection](https://www.mindstick.com/blogs/45023b9e-79fe-4f79-ab49-bbbba6e38981/images/49a36864-8f90-4d94-b1b2-cbb413b3ad93.png)\ \ \ - Google has added unforeseen external reply warnings to Gmail - Google agrees that about 50% - 70% of mails are spam on Gmail - New updates seem to be focused on enterprise customers This isn’t any secret that **phishing attacks** are on the hike and increasingly after moving towards making its Gmail client safer across all platforms for quite some time, Google has now taken another significant step in the same direction. **Google** has announced that the company has now integrated AI through machine learning for early detection of phishing attacks through Gmail. The search giant also claims to have integrated warnings at click-time for malicious links, unintended external reply warnings, and built-in defenses against new threats, apart from machine learning. The latest additions from the search giant seem to be focused on **businesses and Google** says that new security features will help them stay ahead of potential threats. Google said about the machine learning, "Artificial Intelligence and Machine learning helps Gmail block sneaky spam and phishing messages from showing up in your inbox with an accuracy up to 99.9 percent." The search giant has termed this as a crucial factor as according to its own estimates, 50-70 percent of messages that Gmail receives are spam. A dedicated **machine learning** model, meant for early phishing detection, selectively delays messages to perform rigorous phishing analysis and further protect user data from compromise. If you are concerned about the delay, there's no need as the search giant says it happens for less than 0.05 percent of all messages on average, Google stated. In order to find and flag suspicious URLs, The detection modules are integrated with Google Safe Browsing technology that is owned by a California-based company. Google said in its blog post, "These new models combine a variety of techniques such as reputation and similarity analysis on URLs, allowing us to generate new URL click-time warnings for phishing and malware links. Whenever, any new patterns are found, our models adapt more quickly than manual systems ever could, and get better with the time." In order to increase the security awareness at enterprise level, Google has added a quick warning for employees that shows up to confirm if they intended to send an email whenever they are about to send someone outside their company's domain. The company says that to avoid unnecessary warnings, the contextual intelligence through machine learning allows Gmail to understand if the recipient is an existing contact or someone you converse on a regular basis. "With new built-in defenses against **[ransomware](https://www.mindstick.com/blog/11423/how-to-protect-yourself-from-wannacry-ransomware-or-any-malware)** and polymorphic [malware](https://www.mindstick.com/articles/12717/malware-ransom-ware-and-spyware-and-its-effects), Gmail [now](https://yourviews.mindstick.com/view/81402/it-s-liberals-vs-progressives-in-us-politics-now) blocks millions of additional emails that can harm [users](https://www.mindstick.com/news/2244/issue-preventing-users-from-accessing-facebook-s-social-networking-platforms-has-been-resolved)," the company said. [Google](https://www.mindstick.com/articles/43833/google-lighthouse-and-how-is-it-changing-the-way-we-development-and-design-websites) says that it classifies new Gmail [threats](https://www.mindstick.com/blog/12559/real-threats-to-your-business-how-to-prevent-them) by [combining](https://yourviews.mindstick.com/view/85223/hybrid-vehicles-combining-fuel-efficiency-and-performance) "thousands of [spam](https://www.mindstick.com/blog/303939/10-different-ways-to-protect-yourself-from-spam-and-hackers), malware and ransomware signals with attachment heuristics (emails that could be threats based on signals) and sender signatures ([already](https://yourviews.mindstick.com/view/88453/the-hidden-ways-ai-is-already-controlling-your-daily-life) marked malware)." **Read Also: [How to Remove Gmail Account from Android Phone](https://www.mindstick.com/Blog/11440/how-to-remove-gmail-account-from-android-phone)** \ --- Original Source: https://www.mindstick.com/blog/11458/phishing-gmail-gets-ai-integration-for-protection Copyright © MindStick Software Pvt. Ltd. This Markdown version is provided for developers, AI systems, and offline reading.