In-browser antivirus programs aren’t exactly well-loved; many are little more than malware masquerading as legitimate ways to clean systems and proactively detect threats. According to SecurityWeek, however, even well-known antivirus companies sometimes play fast and loose with Chrome extension rules. That’s the problem with AVG’s Web TuneUp, which could expose users’ browsing history to the Internet at large.

Big Business

Chrome extensions are an easy way to get more functionality out of Google’s popular browser. As noted by The Next Web, some are harmless and even hilarious, but the ubiquity of these extensions also makes them popular targets for malware-makers. If packaged with legitimate-seeming software, it’s possible to gain almost unlimited access to victim computers.

Consider the tactic used by some operators of the Fiesta exploit kit earlier this year. As Softpedia explained, when it became clear their payload delivery wasn’t paying off, actors developed the bogus antivirus software “Antivirus Pro 2015,” which always detected threats and then prompted users to purchase the full version.

The solution used to be that users should only download and install legitimate, above-board antivirus software and extensions. But now, a Google security researcher has turned up evidence that antivirus-maker AVG may be putting user privacy at risk with its default Chrome extension.

Bad Practice?

According to Computing, when users install AVG’s security suite, they’re also getting the AVG Web TuneUp extension — whether they like it or not. The problem? It installs a host of JavaScript APIs, some intended to hijack search settings along with the New Tab page. The APIs make it past Chrome’s malware check owing to an extremely complicated installation process, effectively making them a kind of malicious code, one that Google security researcher Tavis Ormandy described as “trash” in an email to AVG.

But installation without permission isn’t the whole story. Ormandy discovered that while most of the APIs were broken, a few created significant flaws in Chrome. For example, a cross-site scripting bug in the “navigate” API made it possible for websites to execute scripts on other domains, such as reading email from mail.google.com or performing other actions at a distance, while AVG’s “recently” API extension exposes user browsing history to the public Internet.

What’s more, Ormandy said that with enough time and effort, attackers could leverage these APIs for use in remote code execution, giving them the ideal hidden malware vehicle: Malicious code hidden inside the offering of a legitimate antivirus company.

Fixing the Chrome Extension

AVG has been working with Google to find a fix, and version 4.2.5.169 should remedy any issues with AVG Web TuneUp — although the Chrome Web Store team has disabled in-line installations of the extension in the meantime. Problematic code aside, there’s a larger issue here: the assumed permission of extension install combined with a clear attempt at API obfuscation.

While Google is investigating possible policy violations, this kind of enforcement is no match for user action. With supposedly above-board companies now willing to play fast and loose with privacy and property, Internet citizens must take a more active role in cataloging, monitoring and removing extensions that reach too far.

More from

Data Privacy: How the Growing Field of Regulations Impacts Businesses

The proposed rules over artificial intelligence (AI) in the European Union (EU) are a harbinger of things to come. Data privacy laws are becoming more complex and growing in number and relevance. So, businesses that seek to become — and stay — compliant must find a solution that can do more than just respond to current challenges. Take a look at upcoming trends when it comes to data privacy regulations and how to follow them. Today's AI Solutions On April…

Why Zero Trust Works When Everything Else Doesn’t

The zero trust security model is proving to be one of the most effective cybersecurity approaches ever conceived. Zero trust — also called zero trust architecture (ZTA), zero trust network architecture (ZTNA) and perimeter-less security — takes a "default deny" security posture. All people and devices must prove explicit permission to use each network resource each time they use that resource. Using microsegmentation and least privileged access principles, zero trust not only prevents breaches but also stymies lateral movement should a breach…

5 Golden Rules of Threat Hunting

When a breach is uncovered, the operational cadence includes threat detection, quarantine and termination. While all stages can occur within the first hour of discovery, in some cases, that's already too late.Security operations center (SOC) teams monitor and hunt new threats continuously. To ward off the most advanced threats, security teams proactively hunt for ones that evade the dashboards of their security solutions.However, advanced threat actors have learned to blend in with their target's environment, remaining unnoticed for prolonged periods. Based…

Third-Party App Stores Could Be a Red Flag for iOS Security

Even Apple can’t escape change forever. The famously restrictive company will allow third-party app stores for iOS devices, along with allowing users to “sideload” software directly. Spurring the move is the European Union’s (EU) Digital Markets Act (DMA), which looks to ensure open markets by reducing the ability of digital “gatekeepers” to restrict content on devices. While this is good news for app creators and end-users, there is a potential red flag: security. Here’s what the compliance-driven change means for…