Skip to main content

Zoom watermarking: pros and cons

Metadata, which gives background information on pieces of data, is typically hidden. It becomes a problem when accidentally revealed. Often tied to photography mishaps, it can be timestamps. It might be location. In some cases, it can be log analysis. Many tutorials exist to strip this information out. This is because it can reveal more than intended when it hits the public domain. Default settings are often to blame. For example, a mobile photography app or camera may embed GPS data by default.

Some people may find this useful; quite a few more may object to it as a creepy privacy invasion.

Well, that's metadata. Now you have an idea what kind of things can lurk without knowledge. We can see what happens when we deliberately enable a data / tagging related function.

Watermarking: what's the deal?

An interesting story has recently emerged on The Intercept, of voluntary data (in the form of watermarks) wrapped into Zoom recordings, which could cause headaches in unexpected ways. Watermarks aren't hidden—they're right there by design, if people choose to use them. And the visual side of this data is supposed to be viewable during the call.

The Intercept talks about accidental identity reveals, via data embedded into calls, in relation to the ever-present videoconferencing tool. You'd be forgiven for thinking the identity reveal referenced in the article had something to do with the watermarks, but no.

The reveal happened because someone recorded a video call and dropped it online, with participant's faces on display. The people involved appear to be at least reasonably well known. The secret identity game was up regardless of what was under the hood.

Cause and effect

What the rest of the article is about, is theorising on the ways embedded metadata could cause issues for participants. Zoom allows for video and audio watermarking, with video of course being visual and so easier to spot. Video displays a portion of a user's email address when someone is sharing their screen. Audio embeds the information of anyone recording the call into the audio, and Zoom lets you know who shared it. You must ask Zoom to do this, and the clip has to be more than 2 minutes in length.

Essentially, video watermarking is to help you know who is sharing and talking during the call. Audio watermarking is to allow you to figure out if someone is sharing without permission. The Intercept explores ways this could cause problems where confidentiality is a concern.

Some identity caveats

If Zoom content is shared online without permission, it may not matter much if revealing metadata is included, unless the video call is audio only. This is because people can be easy to identify visually. Is a public figure of some sort involved? The game is already lost. If they're not normally a public facing persona, people could still find them via reverse image search or other matching tools. And if they can't, a well-known location, or a name-badge, could give them away. There are so many variables at work, only the participants may know for sure.

Hunting the leaker: does it matter?

While the other concern of identifying the leaker is still important, your mileage may vary in terms of how useful it is, versus how much of an inadvertent threat it presents. It's possible the leaker may not care much if they're revealed. They may have used a fake identity, or even compromised a legitimate account in order to do the leaking.

It's also possible that someone with a grudge could leak something then pretend they'd been compromised. If this happened, would you have a way of being able to determine the truth of the matter? Or would you simply take their word for it?

Weighing up the risk

All good questions, and a valuable reminder to consider which videoconferencing tools you want to make use of. For some organisations and individuals, there's a valid use for the metadata dropped into the files. For others, it might be safer on balance to leave them out. It might even be worth using a virtual background instead of something which reveals personal information. It might be worth asking if you even need video at all, depending on sensitivity of call.

The choice, as always, is yours.

The post Zoom watermarking: pros and cons appeared first on Malwarebytes Labs.



from Malwarebytes Labs full article here

Popular posts from this blog

Chaos in a cup: When ransomware creeps into your smart coffee maker

When the fledgling concept of the Internet of Things (IoT) was beginning to excite the world almost a decade ago, perhaps no coffee lover at that time would've imagined including the coffee machine in the roster of internet-connected devices—even in jest. True, the simple, utilitarian coffee machine may not be as popular now as it used to back in the day, but its continued availability within office premises and private home kitchens, plus inherent risks—much like any IoT device—may be in equal footing with your smart speaker , smart doorbell , or smart light bulb . Cybersecurity issues surrounding internet-connected coffee machines are further punctuated by the latest news about how Martin Hron, a reverse engineer from Avast, tinkered his Smarter coffee maker to not only beep and spew out hot water but also deprive you of a nice, morning brew and display a short ransom note. Courtesy of Dan Goodin, Ars Technica Yes, Hron turned his coffee maker into a ransomware mach...

A week in security (December 10 – 16)

Last week on Labs, we took a look at some new Mac malware , a collection of various scraped data dumps , the protection of power grids , and how bad actors are using SMB vulnerabilities .   Other cybersecurity news Millions affected by Facebook photo API bug: An issue granted third-party apps more access to photos than should normally be granted, including images uploaded but not published. (source: Facebook) Bomb threats may be a hoax: An email in circulation urging ransom payments in Bitcoin lest bombs across the US be detonated may well be a fake , according to US law enforcement. (source: The Register) Man jailed for fraud offenses: A man in the UK has been jailed for taking part in fraudulent activities. The main point of interest is surely the spectacular device he built. (source: Met Police) Another Google Plus bug: For six days, developer were able to access profile data not made public by the users. (source: Google) Windows 10 data collection: Reddit use...

Skimmer acts as payment service provider via rogue iframe

Criminals continue to target online stores to steal payment details from unaware customers at a rapid pace. There are many different ways to go about it, from hacking the shopping site itself, to compromising its supply-chain. A number of online merchants externalize the payment process to a payment service provider (PSP) for various reasons, including peace of mind that transactions will be handled securely. Since some stores will not process payments on their own site, one might think that even if they were compromised, attackers wouldn't be able to steal customers' credit card data. But this isn't always true. RiskIQ previously detailed how Magecart's Group 4 was using an overlay technique that would search for the active payment form on the page and replace it with one prepped for skimming. The one we are looking at today adds a bogus iframe that asks unsuspecting customers to enter their credit card information. The irony here is that the s...