You found the perfect image for your blog post, landing page, or ad campaign. It looks great. It fits the vibe. You drag it into your CMS and hit publish.

Six months later, a demand letter lands in your inbox from Getty Images asking for $3,500.

This isn't a hypothetical. It happens to businesses every single day. And the worst part is that it's almost always preventable. If you know how to check whether an image is copyrighted before you use it, you can avoid the legal headaches, the surprise invoices, and the stomach-dropping moment when a law firm's letterhead shows up in your email.

This guide covers seven practical methods to check if an image is copyrighted, from quick manual checks to automated scanning tools that can audit your entire website at once.

Why Checking Image Copyright Actually Matters

Let's get the scary numbers out of the way. Under U.S. copyright law (specifically 17 U.S.C. 504), statutory damages for copyright infringement range from $750 to $30,000 per image. If a court finds the infringement was willful, that ceiling jumps to $150,000 per image. Per. Image.

Most cases never get to court. Instead, copyright holders and stock photo agencies send demand letters, and the amounts in those letters are not small:

Source	Typical Demand Amount
Getty Images	$1,000 - $5,000 per image
Shutterstock	$500 - $3,000 per image
Adobe Stock	$500 - $2,500 per image
Law firms (on behalf of photographers)	$3,000 - $25,000+ per image

A mid-sized e-commerce store with 50 unlicensed images could be looking at $50,000 to $250,000 in potential liability. That's not an edge case. In 2024, Getty Images sent an estimated 100,000+ demand letters worldwide. Shutterstock's enforcement division has been ramping up aggressively since their acquisition of PicScout's detection technology.

And it's not just big companies getting caught. Freelancers, bloggers, nonprofits, and small business owners get these letters all the time. The person who uploaded that image to your WordPress site three years ago might be long gone, but the liability stays with you.

The bottom line: every image on your website is either properly licensed or it's a ticking time bomb. Here's how to figure out which is which.

Method 1: Check Image Metadata (EXIF and IPTC Data)

Every digital image can carry embedded metadata, invisible information baked into the file itself. The two most relevant standards are EXIF (Exchangeable Image File Format) and IPTC (International Press Telecommunications Council).

This metadata can include:

Copyright holder name (the photographer or agency)
Contact information for licensing
Usage rights and license terms
Camera and creation details (date, device, GPS coordinates)
Caption and description fields

If you right-click an image on your desktop and check its properties, you might see some of this data. But most operating systems show only a fraction of what's actually stored in the file. Dedicated tools give you the full picture.

How to do it: Upload the image to a metadata viewer. Look for fields like "Copyright," "Creator," "Rights," "Source," and "Credit." If any of these contain a photographer's name or a stock agency, that image is almost certainly copyrighted and requires a license.

Limitation: Metadata can be stripped. When images pass through social media platforms, CMS systems, or image optimization tools, the metadata often gets removed. A clean metadata result does not mean the image is free to use. It just means the metadata was removed at some point.

Check any image's metadata for free with PixGuard's EXIF/IPTC viewer.

Method 2: Look for Visible and Invisible Watermarks

Visible watermarks are the obvious ones: a semi-transparent logo or text overlay stamped across the image, usually by a stock photo agency. If you see one, the answer is simple. That image is copyrighted, and the watermark is there specifically to prevent unlicensed use.

But visible watermarks are only half the story. Many stock agencies also embed invisible watermarks (sometimes called steganographic or forensic watermarks) into their images. These watermarks survive cropping, resizing, color adjustments, and even screenshots. They are designed to be undetectable to the human eye but readable by automated detection systems.

This is exactly how agencies like Getty track down unlicensed use months or years after the fact. You might think you found a "clean" version of a stock photo somewhere, but the invisible watermark is still there, and their crawlers will find it.

How to do it: For visible watermarks, a visual inspection is enough. For invisible watermarks, you need a detection tool that can analyze the image's pixel data for embedded signals.

Scan an image for hidden watermarks with PixGuard's watermark detector.

Method 3: Use Reverse Image Search

Reverse image search is one of the fastest ways to trace where an image came from and whether it's being sold commercially.

The two main options:

Google Images (images.google.com): Click the camera icon and upload an image or paste a URL. Google will show you visually similar images and the websites where they appear. If the image shows up on Shutterstock, Getty, or a photographer's portfolio, you have your answer.
TinEye (tineye.com): A dedicated reverse image search engine that indexes over 70 billion images. TinEye is particularly useful because it can sort results by "oldest" appearance, helping you identify the original source. It also shows you where the image has been modified or cropped.

How to do it: Upload the image to both Google Images and TinEye. Look for matches on stock photo sites, photographer portfolios, or news agencies. If you find the image being sold on any stock platform, it's copyrighted.

Limitation: Reverse image search works well for popular images, but it's not exhaustive. A photograph taken by a local photographer and never uploaded to a major platform might not show results, even though it's still protected by copyright. Remember that under U.S. and international law, virtually every photograph is copyrighted from the moment it's taken. No registration is required.

Method 4: Search Stock Photo Databases Directly

If you suspect an image might come from a stock library, go straight to the source. The major stock platforms all have search functionality that lets you upload an image or search by keyword:

Platform	Search URL	Notes
Shutterstock	shutterstock.com	Reverse image search available
Getty Images	gettyimages.com	Extensive editorial library
Adobe Stock	stock.adobe.com	Integrated with Creative Cloud
iStock	istockphoto.com	Getty's mid-range offering
Alamy	alamy.com	Strong editorial and news coverage
123RF	123rf.com	Budget stock library

How to do it: Visit each platform and use their search features to look for the image. If you find a match, check the licensing terms. Some images are available under Editorial licenses only (meaning you can't use them commercially), while others require specific Extended licenses for certain use cases.

Limitation: This is time-consuming, especially if you're checking more than a handful of images. And it only covers the major commercial libraries. Smaller agencies, independent photographers, and news organizations have their own catalogs that won't show up in these searches.

Method 5: Compute Perceptual Hashes to Find Near-Matches

This method is more technical, but it's powerful. Perceptual hashing (also called pHash) generates a fingerprint based on the visual content of an image rather than its file data. Two images that look the same to the human eye will produce identical or very similar hashes, even if one has been resized, cropped, slightly color-corrected, or saved in a different format.

This matters because simple file comparison (like checking MD5 or SHA hashes) will fail the moment someone makes even a trivial edit to an image. Perceptual hashing is how the professionals do it.

How it works:

The image is reduced to a small grayscale representation
A discrete cosine transform (DCT) is applied to extract frequency information
The result is encoded as a compact hash string
Two hashes are compared using Hamming distance; a low distance means the images are visually near-identical

If your image's perceptual hash closely matches one in a stock library's database, that's a strong indicator you're dealing with a copyrighted image that originated from that library.

How to do it: Use a perceptual hash tool to generate a hash of your image, then compare it against known databases of copyrighted content.

Generate a perceptual hash for any image with PixGuard's free hash tool.

Method 6: Read the Source Website's Terms of Use

This one sounds obvious, but it's where a huge number of mistakes happen. If you downloaded an image from a website, go back and read that site's terms of use, image licensing page, or creative commons declaration.

Common scenarios that trip people up:

Unsplash and Pexels offer images under permissive licenses, but some images on these platforms have been uploaded without the photographer's permission. If the uploader didn't have the rights, the license they granted you is meaningless.
Wikimedia Commons images are available under various Creative Commons licenses, but many require attribution, and some are non-commercial only. "Creative Commons" does not mean "do whatever you want."
Pinterest, Tumblr, and social media are not sources of free images. The images posted there are almost always owned by someone else, and reposting them on a social platform does not transfer any rights to you.
"Free image" websites with vague or nonexistent licensing pages are often scraping images from other sources. If a site offers millions of "free" images but can't clearly explain where they came from, don't trust it.

How to do it: Find the original source of the image. Read every word of their licensing terms. Look for the specific license type (CC0, CC-BY, CC-BY-SA, Editorial Only, Rights Managed, Royalty Free). If you can't find clear licensing terms, assume the image is copyrighted and not available for your use.

Method 7: Use an Automated Copyright Scanner

All of the methods above work, but they share one problem: they're manual. If you're checking a single image before you use it in a blog post, that's manageable. But if you need to audit an existing website with hundreds or thousands of images, manual checking is not realistic.

This is the problem automated copyright scanners solve. These tools crawl your website, extract every image, and run them through multiple detection methods simultaneously: metadata analysis, watermark detection, perceptual hashing, reverse image matching, and AI-powered visual comparison against databases of known copyrighted content.

What an automated scanner checks for you:

Check	Manual Time	Automated Time
EXIF/IPTC metadata	2-3 min per image	Instant
Visible watermarks	30 sec per image	Instant
Invisible watermarks	Requires specialized tools	Instant
Reverse image search	5-10 min per image	Seconds
Stock database matching	10-15 min per image	Seconds
Perceptual hash comparison	Requires technical setup	Instant

For a website with 500 images, the manual approach would take roughly 40 to 80 hours. An automated scanner does it in minutes.

This is what we built PixGuard to do. It combines CLIP-based visual understanding, perceptual hashing, computer vision watermark detection, and metadata analysis into a single scan. You point it at your website, and it flags every image that might be copyrighted, along with the likely source and the confidence level of the match.

It's not the only option out there, but if you're serious about clearing your website of copyright risk, an automated approach is the only way to do it thoroughly at scale.

What to Do If You Find Copyrighted Images on Your Site

So you've checked, and you've found copyrighted images. Now what? You have three paths:

Remove the Image Immediately

The simplest and safest option. Replace the image with one you've properly licensed, one from a verified CC0 source, or one you created yourself. If a copyright holder hasn't contacted you yet, removing the image significantly reduces your risk, though it doesn't eliminate past liability entirely.

Purchase a Retroactive License

If you want to keep using the image, contact the copyright holder or the stock agency and buy a license. Some agencies offer retroactive licensing at a premium (typically 2x to 5x the normal rate), but it's almost always cheaper than fighting a demand letter. This is especially viable when the image is central to your branding or marketing and replacing it would be costly.

Respond to a Demand Letter (If You've Already Received One)

If you've already received a demand letter, do not ignore it. Ignoring demand letters is how $1,500 settlements turn into $15,000 lawsuits. Respond promptly, consider hiring a copyright attorney for amounts over $5,000, and know that the initial demand is almost always negotiable.

For a more detailed breakdown of handling demand letters, read our guide on what to do when you get a stock photo copyright claim.

Common Myths About Image Copyright

These misconceptions get people into trouble constantly. Let's clear them up.

"I found it on Google, so it's free to use." Google is a search engine, not a licensing platform. Every image in Google's results is owned by someone. Google Image Search even displays a "Licensable" badge on images it detects as commercially licensed. Finding an image on Google gives you zero rights to use it.

"I credited the photographer, so it's fair use." Attribution is not a license. Giving credit is polite, and some Creative Commons licenses require it, but it doesn't substitute for a license where one is needed. Fair use is a specific legal defense with four factors that courts evaluate on a case-by-case basis. "I gave credit" is not one of them.

"I edited the image, so it's a new work." Cropping, adding filters, flipping, resizing, overlaying text, or running it through an AI style transfer does not create a new, independent work. Derivative works still require permission from the original copyright holder. Stock agencies' detection tools are specifically designed to catch edited versions.

"The image didn't have a copyright notice, so it's public domain." Copyright protection is automatic. Under the Berne Convention (which covers 181 countries), a photograph is copyrighted from the moment it's taken. No notice, no registration, and no watermark is required. The absence of a copyright symbol means nothing.

"It's on social media, so it's public." Posting an image on Instagram, Twitter, or Facebook does not put it in the public domain. The photographer retains copyright. Social media platforms' terms of service grant the platform certain usage rights, not you.

"My web developer / freelancer handled the images, so it's their problem." The website owner is liable for the content on their site. If your developer used unlicensed images, the demand letter comes to you. You might have a separate claim against the developer, but that's a different legal battle.

A Quick Decision Framework

When you come across an image you want to use, run through this checklist:

Can you identify the photographer or copyright holder? If yes, contact them for a license.
Is the image on a stock photo platform? If yes, buy the appropriate license.
Does the source provide a clear, verifiable license (CC0, CC-BY, etc.)? If yes, follow the license terms exactly.
Is the image listed as public domain on a credible source (e.g., Library of Congress, NASA)? If yes, you're clear.
Are you unsure about any of the above? If yes, don't use the image.

When in doubt, the safest image is one you took yourself, one you commissioned, or one you licensed through a reputable platform with a clear paper trail.

Protect Your Website Before a Demand Letter Does It for You

Checking one image at a time works when you're publishing new content. But what about the hundreds or thousands of images already on your site? The ones uploaded by former employees, contractors, or that intern from 2022?

That's the gap most businesses don't think about until it's too late.

PixGuard scans your entire website and flags potentially copyrighted images before a stock agency does. Every new account gets 100 free credits, enough to audit a small to mid-sized site. No credit card required.

Scan your website for free with PixGuard -- 100 credits included.

How to Check If an Image Is Copyrighted (2026 Guide)