> IT-Sentinel.com

// Cybersecurity & IT News Aggregator - Real-time Threat Intelligence Feed

NEWS CVE
← messages.back_to_articles

> Raven AI lets cops find your home from a photo taken inside

[SOURCE] Malwarebytes [DATE: 15/10/2026 11:09] [LANGUAGE: EN]
Photos you take with your phone and some digital cameras can contain hidden information, including where they were taken. You can strip that data out before sharing a photo with someone, which would help to protect your location. But the image itself may still give you away. A company called Graylark says its Raven AI tool can match an indoor photo to a specific property using visual clues alone. No GPS data or view out of the window required.We’ve seen this company before. Its tool, GeoSpy, could work out where a photo was taken by analyzing features in the image, such as vegetation, buildings, and other landmarks.Graylark itself grew out of experiments founder Daniel Heinen and his brothers carried out after reading about using AI to play GeoGuessr, a geography game in which people look at street-level images and try to work out where they were taken.The company has since taken it further. According to 404 Media, Graylark has renamed the tool as Raven and expanded it to identify indoor locations as well. It uses visual clues in the image, without needing location metadata.The tool works by matching images against property photos. An image of something as banal as a washer/dryer sitting in a nook can return a list of ranked results. The tool’s Short-Term Rental ID feature can search what Graylark describes as millions of rental listings worldwide, using an interior photo or a frame from a video.However, the system targets more than just rental properties. Its Find Property feature searches for properties using exterior photos and, now, interior ones in supported cities.This is more than just reverse image search, which can help you see if the same photo you’re looking at has been used elsewhere. The service can match different views of the same property by identifying features ranging from the shapes of wall tiles to the structure of a room and the color of its surfaces.Law enforcement is already using the tool, for which founder Daniel Heinen landed $10.7m in seed funding in August. 404 Media said that Miami-Dade Sheriff’s Office and the Los Angeles Police Department bought access to the product.According to Heinen’s social media posts, Raven can now tell which hotel you stayed in from a single photo. He cited a feature called HotelID that isn’t yet listed on the website. The Raven website also cites other uses. For example, it can tell you a car’s exact make, model, and year from a picture of its interior.The tool works from identifiers such as email addresses, too. Investigators can find public photos on linked accounts and then use those photos to look for location clues. Graylark says this does not give investigators access to private accounts or messages.Heinen says the short-term rental feature uses images from millions of rental listings around the world. According to 404 Media, Airbnb said it had no relationship with Graylark, and Zillow said scraping was against its terms of service. Those statements don’t confirm how Graylark obtained its imagery.Heinen’s ambitions don’t seem to stop here. On September 8, he mentioned that he was experimenting with recent footage from tens of thousands of dashcams.He has said that this kind of system can help combat human trafficking. A George Mason University study found that in around 80% of federal sex trafficking cases active in 2018 and 2019, victims were exploited at hotels or motels. And perhaps AI can help with unsolved crimes if used properly. Any suggested location still needs to be checked against other evidence.But technology can also be misused, particularly when it comes to stalking. Heinen said on X that Raven is only available to enterprises, government, and law enforcement agencies. Other websites and consumer apps use the GeoSpy name, but Graylark says third-party products using that name are not operated by or affiliated with the company.This brings us to what you can do to protect yourself should someone use this kind of tool for nefarious purposes.How to keep your home more privateThink before publishing interior or exterior pictures of your home online. Even Google Photos can estimate where a photo was taken using landmarks and similarities with your other photos that have known locations, it says.Check the frame for distinctive features before posting. Look for distinctive tiles, room layouts, views through windows, house numbers, or mail showing your address. Crop these out where practical. Limit who can see your photos. Review the audience for your social posts and photo albums, including older ones. When traveling, consider waiting until you’ve left before posting recognizable accommodation photos.Review old property listings. Ask your agent and the relevant property websites whether they can remove interior photos from old listings. Removal cannot recall copies already downloaded or collected. If you’re advertising a rental, use approximate-location settings where available, but remember that the listing photos themselves may still identify the property.Remove location metadata before sharing. This won’t prevent visual matching, but it removes another way to find where a photo was taken. Don’t assume every service removes it for you.Apple users can do this directly on iOS: When sharing from Photos, tap Options and turn off Location.Google Photos lets you remove estimated or manually added locations, but not those recorded by your camera. Check location-sharing settings for each album, link, or conversation. Photos downloaded and shared outside Google Photos may still contain the original location data.Samsung users can choose to remove location data when sharing from Samsung Gallery, where the option is available.Browse like no one’s watching. Malwarebytes Privacy VPN encrypts your connection and never logs what you do, so the next story you read doesn’t have to feel personal. Try it free → 
[messages.read_original_source] →