Meta explained its AI privacy policies for smart glasses to me. I’m still worried

I’m wearing Meta Ray Ban It turns on and off when I travel to take photos, make phone calls, and listen to music. Technology is great, fun and convenient.

I also knew that Meta’s privacy policies could be a concern, but now I’m more concerned about them than ever.

My fears increased after a number of friends and colleagues A report has been shared About the ability of external Meta contractors in Kenya to view sensitive information such as images of bank records, nudity and sexual encounters recorded on Meta glasses (resulting in Class action lawsuit).

What limits does Meta set to protect people’s privacy? I’m done with it Meta Terms of Service Online and in the Meta AI app, but that wasn’t helpful.

I wanted some answers. So I contacted Meta’s communications team to get clarity.

But even after getting the official response from Meta about where to draw the lines, I’m still frustrated and unsure. While many people are rightfully concerned about someone And secretly record it with smart glasses,There is also another problem: When are these glasses likely to share what you have recorded with others?

Here’s the short answer: Does Meta Glasses have outside contractors potentially accessing your data? Yes, sometimes – if you use AI services. If you’re not using those AI services, according to Meta, you should be fine. But even then, I don’t know where the “AI services” wall will be clearly drawn. This is one of my biggest fears.

Meta has a long history of privacy and trust issues, extending back into the last decade and into the 20th century Cambridge Analytica scandal. These issues did not come up Meta virtual reality headsetswhich doesn’t have many data-gathering AI services, but the company’s smart glasses do. These services will continue to grow and become more capable over the next few years. Meta’s iconic Ray-Ban sunglasses – more than 7 million pairs Sold last year – they are the leaders Full wave Camera-enabled artificial intelligence glasses and wearable devices Coming from a number of companies, with Google enters the mix Later this year.

If you’re interested in the Meta glasses, which are, as a technical achievement, the best quality camera with audio Smart glasses For now, you should keep these concerns in mind. As smart glasses become always-on AI devices, we’ll face more questions about how comfortable you might feel relying on their services — and what all cloud-based AI technology companies need to do to make those policies clearer.

Below, I’ll share Meta’s responses at length so you can understand my reasons – as well as make your own assessment of the risks.

Using AI services with Meta Ray-Bans

If you use artificial intelligence — for example, to analyze something you see or get a translation — third-party contractors may see what you record.

Here’s what the company told me: “Ray-Ban Meta glasses help you use artificial intelligence, hands-free, to answer questions about the world around you. Unless users choose to share the media they’ve captured with Meta or others, that media stays on the user’s device.”

But then there’s this: “When people share content using Meta AI, we sometimes use contractors to review this data for the purpose of improving people’s experience, as many other companies do. We take steps to filter this data to protect people’s privacy and to help prevent review of identifying information.”

The assumption you can make from this is that any time you use Meta’s AI Services, Meta may be using outside contractors to review the information.

While Meta promised to properly filter information to remove sensitive data or details, this alarming news report stated that contractors in Kenya were annotating screenshots taken from glasses that contained sensitive images that were clearly visible.

This has me particularly concerned about what happens when people use Meta AI for ancillary purposes: that is, as a way to “see” when you can’t with your own eyes. Would it be risky to look at and read personal documents again? Since Meta hasn’t properly introduced any kind of special, hard-coded AI features on its glasses, this may be the case.

Meta says this regarding privacy protection: “We have strict policies and guardrails that intentionally limit what information contractors see.”

But again, I don’t actually know what those strict policies or guardrails are.

“We take steps to filter this data to protect people’s privacy and to help prevent review of identifying information,” Meta added.

This doesn’t help clarify any of the details. I’m going to rely on trust here, and that’s not ideal at all.

I have to assume that anything done via cloud AI services, such as using Meta, can be seen to some extent by outside contractors. And you should too.

Capture photos and videos with Meta’s Ray-Bans

Meta glasses don’t use AI all the time, and I don’t either. In fact, I mostly use the Meta glasses for recording photos and videos, listening to music, and making phone calls. I don’t use AI much, partly because Meta’s AI doesn’t interact much with or control my other personal data or even my iPhone.

For non-AI photo and video recording, things should be safe… I think.

I asked members of the communications team whether photo or video recordings I created with the glasses, which were not included in the AI-based recalls, could be subject to a third-party contractor’s bid. They said this: “To be clear, the photos and videos users take with their AI glasses that are simply stored in their phone’s camera roll are… no Meta uses it to develop and improve artificial intelligence. If you just record a video or take a photo using the camera button on the glasses, that media stays on your phone. Unless you choose to share media you captured using Meta or otherwise, that media will remain on your device.”

That looked promising. But with the Meta goggles settings, storage gets a little cloudy… literally. In the Meta AI app’s glasses privacy settings, Cloud Media toggles the prompts to “allow your photos and videos to be sent to the Meta Cloud for processing and caching.”

Did cloud media mean that my personal photos and videos were open to potential annotations by an outside contractor? According to Meta, no. According to Meta, any commands that use AI to send images or use auto-capture modes enabled by the toggle on Cloud Media will also be safe.

As the company puts it: “Some features, such as sharing from your glasses using your voice (“Hey Meta, send a photo”), seamless automatic media import, or Auto Capture, where the camera automatically captures photos or videos when you start the feature (useful for moments when you might want to capture content without manually launching the camera via a button or voice), may require photos and videos to be sent to the Meta cloud for processing and temporary storage. If you sign up for cloud media services, photos and videos sent from frames or automatically imported To your phone is not subject to human feedback Enabling cloud media services is optional and not turned on by default.

The Meta doesn’t clearly define exactly what “Cloud Media” is, other than a place to temporarily store your photos and videos so they can be manipulated with voice commands. What concerns me is how a wall is drawn around “private” media versus “AI-connected” media. It makes me want to turn off Cloud Media, which means photos and videos are only stored in my phone’s photo library.

What to do about AI glasses now?

I still love the camera and audio features of the smart glasses and am intrigued by the upcoming AI features. But I’m also deeply concerned about the uncertainty of where the line is drawn between what might be explained by a third party, and what might remain private. Meta uses these third parties to help train the AI, or perhaps to moderate content. It’s a reminder of how cloud-dependent and out of our control many AI services are.

I’m even more concerned when I think about the reports of Meta wanting it Add facial recognition And more to her smart glasses.

Meanwhile, more AI glasses and AI wearables with cameras are coming as well. Google is next. And all of these companies need to explain how they use data from these devices, how they protect our privacy issues, and how we users can manage them — if that’s possible at all. It’s not at all easy to understand how Meta Glasses handles AI data, or where it’s sent. I hope this story helps you better understand where the lines are.

However, I have to admit that I’m less likely to use the Meta glasses for anything personal or data-sensitive. Vacation glasses? A tool for quick social snapshots of work. Do I stream anyway? Experiments with artificial intelligence? I think so.

But if Meta aims to be a very helpful tool for us via wearable AI devices, and it doesn’t want everyone to do that Describing them as “deviant glasses”, As with people already, better work needs to be done, quickly.