In September 2026, the screenshot sharing service Gyazo disclosed a breach affecting approximately 23.62 million user records. The compromised data included email addresses, password hashes, authentication tokens, and image metadata. The company confirmed no payment information was exposed. It also confirmed that, as of the disclosure date, it had no evidence of misuse.
What the announcement did not dwell on — but what is worth examining — is what that image metadata represents, and what the contents of 23 million users’ screenshot libraries would tell anyone with access to them.
Screenshots are a category of personal file that most people handle without much thought. You take one to save a confirmation number, capture a conversation, preserve a receipt, or record something on a screen before it disappears. Then the screenshot lives in your photo library, or in a screenshot-sharing service, essentially forever — because screenshots feel ephemeral even as they accumulate into a detailed record of your digital life.
What a Screenshot Actually Contains
A screenshot is more than a picture. It is a snapshot of a moment in your digital life, and the contents of that snapshot often include information that would surprise you if you audited it.
Financial information. Bank balance notifications, transfer confirmations, invoice totals, subscription charges, investment portfolio values. If you have ever screenshotted a payment confirmation or a bank statement, that information exists as an image in your library.
Health information. Symptom searches, appointment confirmations, medication reminders, fitness metrics, mental health app data. Health-adjacent information shows up in screenshots because people capture it for reference — before a doctor’s appointment, to remember a medication dose, to share with a family member.
Communications. Screenshotted text conversations contain the contents of private exchanges. The sender and recipient do not need to be named in the screenshot for the content to be sensitive — the text itself may be.
Account credentials and tokens. People screenshot setup screens, QR codes for multi-factor authentication, temporary passwords sent via email, and setup flows that display credentials. These screenshots represent live credentials if the information in them has not been rotated since the screenshot was taken.
Location information. Screenshots of maps, ride confirmations, delivery tracking, location-tagged social media posts, and navigation directions disclose specific addresses — home, work, family, medical facilities.
Biometric and identity information. Passport photos, ID scans, and face-visible screenshots from video calls are common contents of personal screenshot libraries.
What Image Metadata Reveals
The Gyazo breach included image metadata — information about the screenshots themselves rather than their contents. This includes file creation timestamps, user identifiers, and potentially device information, depending on what the service captured and stored.
Metadata about screenshots is worth considering separately from metadata about photographs taken with a camera. A photograph’s metadata includes GPS coordinates from where it was taken, camera model, and time of capture. A screenshot’s metadata includes the time the screen was captured and, often, the device and operating system it was captured on.
Timestamps from a large screenshot library create a timeline of your device usage patterns. Combined with the content of screenshots — a bank notification at 8:43 AM, a message exchange at 11:17 PM — they reveal behavioral patterns that go beyond what the individual screenshots contain.
Authentication tokens included in the breach data are a more immediate concern. A session token for a web service allows the holder to authenticate as you without a password. When tokens are compromised in a breach, the first-order concern is unauthorized account access to the services those tokens belong to — though Gyazo indicated no evidence of misuse, the exposure window between the breach and the disclosure is a variable that users cannot independently assess.
How Screenshots Accumulate Without Review
Part of what makes screenshots a privacy blind spot is how they are created and stored. Unlike documents you create intentionally — a journal entry, a note, a saved file — screenshots happen in a moment of convenience and immediately recede from active attention. You needed that confirmation number an hour ago. You no longer think about the screenshot that captured it.
Cloud photo services that automatically back up your camera roll include screenshots in the backup, by default. This means every screenshot you have ever taken on your phone has likely been uploaded to Google Photos, iCloud, or your backup service of choice — where it is stored, indexed, analyzed, and subject to that provider’s terms of service.
Screenshot-sharing services like Gyazo exist because people want to share a captured screen quickly. The convenience model — one click, a shareable link — means the screenshots accumulate in the service’s storage continuously, often without users returning to manage or delete them. An account opened years ago may contain thousands of screenshots the user has entirely forgotten.
The result is that many people have a large, largely unreviewed archive of their screen activity stored in one or more cloud services, with no active sense of what that archive contains or what it reveals.
The Specific Risk of Screenshot-Sharing Services
General cloud storage services are a known risk that receives some attention. Screenshot-sharing services are a less-examined category, and the risk profile has a few distinctive features.
Content is created under lower scrutiny. When you save a document to cloud storage, you typically know you are creating a persistent file. Screenshots are taken in a moment of convenience, without the deliberate framing that comes with intentional file storage.
Sharing links persist. Many screenshot-sharing services generate a URL for each uploaded screenshot. The user shares the link and moves on. The link continues to work unless the user returns to delete the screenshot — which most users do not, because the screenshot feels like it is already in the past.
Services in this category may have smaller security teams. A dedicated screenshot-sharing service is a more focused, often smaller product than a major cloud storage platform. The security investment may not be proportionate to the data being stored.
Deletion expectations are unclear. When you share a screenshot via a link and the recipient views it, the reasonable expectation is often that the sharing is complete. The actual situation — the screenshot persists in the service’s storage indefinitely — is often not how users think about it.
What to Do With Your Screenshot Library
A few concrete steps for treating screenshots more deliberately:
Audit what is in your screenshot library. Open your phone’s photo library and filter to screenshots. Scroll through what you find. You will likely be surprised by the range of sensitive information that has accumulated.
Enable screenshot categorization on your phone. Both iOS and Android allow screenshots to be viewed as a separate category in the native photos app. This makes it easier to review and delete them in bulk.
Turn off cloud backup for screenshots specifically. On iPhone, you can exclude screenshots from iCloud syncing using the Hidden album, though the integration is imperfect. Third-party apps that offer selective backup categories may be more reliable. On Android, Google Photos allows you to back up specific folders; removing the Screenshots folder from backup stops future uploads but does not delete historical ones.
Review and delete from screenshot-sharing services. If you use a service like Gyazo, Droplr, or similar, log in and review what is stored there. These services often have bulk delete options that are not prominently surfaced but do exist.
Rotate credentials that may have been screenshotted. If you have ever taken a screenshot of a password, a two-factor authentication backup code, an API key, or a setup QR code, those credentials should be treated as potentially compromised if the screenshots are in any cloud service. Change them.
The Storage Problem Screenshots Represent
The Gyazo breach illustrates something worth generalizing beyond this specific incident: sensitive personal information ends up in services that were not designed as secure document storage, through behaviors that feel casual and inconsequential.
A screenshot of your bank balance is just a convenience — taken to remember a number, forgotten immediately. It is also a financial document, now in a cloud service, potentially for years.
The gap between how screenshots feel and what they actually are — persistent, detailed records of your screen activity containing potentially sensitive information — is the root of the problem. Bridging that gap means treating the screenshot library the way you would treat any sensitive archive: knowing what is in it, knowing where it is stored, and making deliberate choices about what persists.
If you store personal screenshots alongside documents, invoices, and records you are intentional about, in a service built around privacy rather than sharing convenience, you gain the same visibility and control over all of it. daftei stores any file you choose to save — screenshots included — with TLS 1.3 in transit and AES-256 at rest, without sharing links that persist by default, without AI analysis of your content, and without using what you store to improve third-party models.
The practical question after a breach like Gyazo’s is not just whether your account was in the affected 23.62 million. It is whether the services where your screenshots currently live meet a standard you have actually thought about.