Google Chrome is getting new defenses against online scams that trick users into visiting dangerous websites or running malicious commands on their own computers.
Google announced new Gemini-powered scam detection and warnings on Oct. 20, targeting deceptive websites and ClickFix attacks that manipulate victims into executing malicious commands. The new defenses could benefit more than 1 billion Chrome users who have enabled Enhanced Protection, according to Google's previously reported adoption figures.
For Chrome users and IT administrators, the changes address a growing security problem: Attackers increasingly rely on convincing people to compromise their own devices rather than exploiting software vulnerabilities directly.
Chrome uses Gemini AI to identify suspicious websites
Google detailed the new capabilities in its Oct. 20 Chrome security announcement, explaining how artificial intelligence is being incorporated into Safe Browsing protections.
The company previously introduced on-device Gemini scam detection in 2025 to help identify technical support scams. That capability has now moved to a more powerful server-side Gemini model, expanding coverage to all users who enable Enhanced Protection.
The updated system is designed to recognize a broader range of fraudulent websites, including fake antivirus warnings and deceptive browser-update pages.
According to Google, Gemini evaluates suspicious page content using several indicators:
- Brand representation: Whether a website appears to impersonate a legitimate company or service.
- Page intent: Whether the content suggests an attempt to deceive visitors.
- Scam likelihood: An AI-generated assessment that contributes to the browser's overall security verdict.
Chrome combines these findings with domain reputation and other Safe Browsing signals to determine whether a website poses a threat.
Google says page text is sent for AI analysis only when a website exhibits high-confidence indicators of a scam. The company also says any content sent to its servers is processed temporarily and is not logged, retained, or used to train AI models.
The changes arrive as Chrome continues strengthening its defenses against both software vulnerabilities and malicious websites.
Earlier this month, Google released Chrome 155 with fixes for 247 security vulnerabilities, including four critical flaws.
The latest announcement addresses a different attack surface: Websites that persuade users to perform dangerous actions themselves.
New Chrome warnings target ClickFix malware attacks
One of the most significant additions targets ClickFix, a social engineering technique that uses fake technical problems to convince victims to execute malicious commands. These attacks typically begin when a user visits a compromised or fraudulent website displaying a fabricated error message.
The page might claim that the browser requires an update, a security check has failed, or a technical problem needs immediate attention. Victims are then instructed to copy a command and execute it using a legitimate operating-system utility.
On Windows, these instructions frequently involve tools such as PowerShell or Command Prompt. Because these utilities are legitimate parts of the operating system, attackers can exploit them to download malware or execute malicious scripts while avoiding some conventional browser defenses.
For users with Enhanced Protection enabled, Chrome's new protection is designed to identify suspicious commands when they are copied from websites.
If the browser detects a potential ClickFix script, Google says the clipboard payload is stripped of personally identifiable information before being submitted to Safe Browsing for further analysis.
When the system determines that the command is dangerous, Chrome displays a warning. This intervention happens before the user executes the copied command, creating an additional opportunity to prevent compromise.
The protection does not mean Chrome monitors every command executed on a computer. It specifically addresses suspicious copying activity associated with potential ClickFix attacks.
The threat is already well documented. Previous eSecurity Planet reporting identified more than 250 ClickFix domains used to conceal macOS malware, demonstrating how attackers have adapted the technique to target different operating systems.
Google's new warnings aim to disrupt this type of attack before users move from a deceptive webpage to executing malicious instructions.
Google introduces separate warnings for suspicious and dangerous websites
Google is also changing how Chrome communicates potential threats from websites. Previously, users were familiar with the browser's prominent red warning pages for known dangerous destinations.
Chrome now distinguishes between suspicious and dangerous websites when Enhanced Protection is enabled:
- Suspicious website warnings appear when a page displays deceptive characteristics associated with malicious websites, but the system has lower confidence that the site poses a threat.
- Dangerous website warnings continue to appear when Chrome has stronger evidence that a website is malicious and presents a high risk of harm.
Google says the two warning categories use different visual designs, colors, icons, and messaging to help users understand the level of risk. The distinction also allows Chrome to warn users about a broader range of potentially deceptive websites without treating every suspicious page as a confirmed threat.
As SecurityWeek reported on Google's new Chrome defenses, the company is combining AI-based website analysis with more targeted warnings against social engineering. The approach could be particularly useful against newly created scam websites that have not yet accumulated enough malicious activity to appear in conventional threat databases.
However, the new warnings are not proof that a website is malicious, and users should not assume that a page is safe simply because Chrome does not display an alert.
What Chrome users and security teams should do now
The new protections are tied to Chrome's Enhanced Protection mode, which provides additional security checks beyond the browser's Standard Protection setting.
Google reported in 2025 that more than 1 billion Chrome users had enabled Enhanced Protection. The company has not provided a newer exact figure in its latest announcement.
Google says the server-side Gemini capability expands scam detection to all users who have enabled Enhanced Protection, rather than limiting it to devices supported by the earlier on-device system.
Users can review their settings by opening Chrome and navigating to Settings > Privacy and security > Security. From there, they can review Safe Browsing options and choose whether to enable Enhanced Protection.
Google says Enhanced Protection users receive additional safeguards against phishing, scams, and other online threats, although the setting also involves additional security-related data sharing.
Users should review those privacy considerations before changing their settings.
- For organizations managing Chrome across employee devices, the changes offer another opportunity to reduce exposure to social engineering attacks.
- Security administrators should ensure that browsers remain up to date, review managed Safe Browsing policies, and reinforce employee training on suspicious websites.
- In particular, employees should be instructed never to execute unfamiliar commands simply because a website claims that a browser update, security verification, or technical repair is required.
Organizations should also maintain endpoint protections capable of detecting suspicious PowerShell activity and other potentially malicious command execution.
The importance of layered defenses became apparent in September, when four espionage groups used the BlueMoon exploit kit to target Chrome and Windows vulnerabilities.
Although that campaign involved software exploitation rather than ClickFix-style social engineering, it demonstrated why organizations need multiple layers of browser and endpoint security.
Why Google's new protections matter for browser security
Google's latest changes reflect a broader shift in how attackers attempt to compromise computers.
Traditional browser security has focused heavily on identifying malicious downloads, blocking dangerous websites, and patching exploitable software vulnerabilities. ClickFix attacks challenge those defenses by persuading users to perform malicious actions using legitimate operating-system tools.
A website does not necessarily need to exploit a browser vulnerability if it can convince someone to voluntarily execute a harmful command. By introducing Gemini-powered scam detection and warnings for suspicious clipboard activity, Google is attempting to interrupt that process before the attacker gains a foothold.
The protections have limitations. AI-based detection can miss new techniques, and users can still disregard security warnings or encounter malicious instructions through other channels.
For security teams, the most important takeaway is that browser security must extend beyond patch management. Keeping Chrome updated remains essential, but organizations also need to address the human decisions attackers exploit.
For individual users, the advice is straightforward: Enable appropriate browser protections, remain skeptical of unexpected technical warnings, and never run unfamiliar commands copied from websites.
Google's new defenses may help detect more deceptive activity, but recognizing suspicious instructions remains essential to staying safe online.
More cybersecurity news: For another example of how attackers exploit browser trust, read how a hijacked Chrome extension was used to deliver ClickFix malware and steal sensitive data.