You finish a post and need a picture for it. You search “free images,” a dozen sites come up, and grabbing one seems harmless. Years later some people get a letter from a law firm about that picture. Free does not mean the copyright went away. Free means someone gave you permission on certain terms, and the terms differ from site to site. That gap is where the trouble starts.
Getty is paid. But three free doors have Getty’s name on them
Start with the biggest name, because it confuses people the most. Getty Images is a paid stock photo company. You cannot download a photo there and drop it into your blog. Yet there are three routes to free imagery connected to Getty, and mixing them up is easy.
1. Getty owns Unsplash
The largest free photo site on the web is Unsplash. Getty Images acquired it in March 2021. At the announcement Getty said Unsplash would keep running as its own brand and that licenses on photos already published could not be changed retroactively. It still runs free today. If you are looking for free photos, this is your closest point of contact with Getty.
2. The embed tool on gettyimages.com
Open a photo on Getty’s site and you will find an embed icon. Click it and you get an iframe snippet. Paste that into your post and the photo appears, with the photographer’s credit and a link back baked in. You are not downloading anything — Getty keeps the file on its servers and you get a window onto it.
The catch is the use. Embeds are for editorial, non-commercial purposes, meaning things that are newsworthy or of public interest. No ads, no promotions, no merchandise, and nothing that suggests the people in the photo endorse your product. Having ads on your site does not by itself make the use commercial — news sites run ads and still qualify — but an image inside an ad does.
3. Regional Getty sites with their own free sections
Some Getty-affiliated regional sites run a free image section with their own rules, usually behind a free account and with a download quota. Those terms are set locally and can differ from what gettyimages.com says. Read the page you are actually downloading from.
The big three free photo sites already have owners
Unsplash, Pexels, and Pixabay carry most of the traffic. All three started independent. None of them is now.
| Site | Owner | Attribution | Note |
|---|---|---|---|
| Unsplash | Getty Images (2021) | Not required | Paid tier Unsplash+ exists alongside |
| Pexels | Canva | Not required | Photos and video |
| Pixabay | Canva | Not required | Own license since January 2019 |
Pixabay deserves a footnote. It used to release everything under CC0, which is as close to public domain as a license gets. That changed on 9 January 2019, when it switched to its own Content License. Anything published before that date and anything after it sit under different terms. Old blog posts still tell you “Pixabay is CC0, do whatever you want.” Do not run on that.
Unsplash never says CC0 either. Read its license page and you will find a custom set of terms instead.
The lines all three draw
Put the license texts side by side and the same prohibitions show up. Worth memorizing once.
- Selling unaltered copies. Posters and prints are the classic example. Unsplash puts it as “images cannot be sold without significant modification.”
- Redistributing or selling the files on another stock photo or wallpaper site.
- Collecting the images to build a similar or competing service. Unsplash names this one directly.
- Implying that a person or brand in the photo endorses your product.
- Using a photo as part of your trademark, design mark, trade name, or service mark.
- Showing identifiable people in a bad light or in an offensive way.
Dropping a photo into a blog post trips none of these. Designing a logo or printing merchandise is a different conversation.
Icons are where attribution bites
More people get caught by icons than by photos, because free icons often come with a credit requirement attached.
Flaticon’s free tier requires attribution. You have to credit the author with a visible link on every page where the icon appears, and only a Premium subscription removes that. Icons8 works the same way — free means credit, paid means no credit.
If you would rather not manage credit lines, go open source. Lucide ships under the ISC license and Bootstrap Icons under MIT. Both allow commercial use with no visible attribution; you keep the license file in your source and you are done. For anything on a website, that is where I pick from. Chasing a prettier icon and inheriting a credit obligation is a bad trade.
Fonts
Google Fonts is the easy answer. Everything there is distributed under open source licenses, commercial use included, and you can either self-host the files or link them. No credit line needed.
Where people slip is fonts found elsewhere. A directory that lists “free fonts” is usually an index, not the rights holder, and each font carries its own terms from whoever made it. Web use, print, printing on products you sell, and embedding the font file in an app are four separate permissions, and a font can allow one while blocking another. Open the license summary before you commit a font to a design you will reuse.
Design tools: watch for the crown
If you are making the image rather than finding it, Canva is the default. The free plan covers commercial use, and Canva says attribution is not required for its free photos, music, and video.
Two things to notice. Templates mix free and premium elements, so a free template can still contain a paid photo that stamps a watermark on your export. And Canva’s own terms say that content sourced from Pixabay or Pexels stays under those sites’ licenses. Being inside the tool does not strip the terms off the file.
If you run WordPress, it is already in your editor
This one gets overlooked. The WordPress block editor has Openverse built in, so you can search and insert an image without leaving the post. Openverse is run by WordPress.org and searches more than 800 million openly licensed and public domain works from across the web. It started life as CC Search at Creative Commons and moved to WordPress in 2021.
Read the warning Openverse puts on itself, though: it cannot guarantee that the license information is accurate and tells you to verify before using a work. It indexes metadata from other people’s sites, so errors upstream land in the results. Search there, then click through to the source and check the license yourself.
Public domain collections are the quiet option
Wikimedia Commons holds tens of millions of freely licensed files, and several museums have opened their collections outright — the Metropolitan Museum of Art and the Smithsonian both publish large public domain sets. NASA’s image library is another, though a few items there carry use restrictions, so read the page rather than assuming. Copyright expires after the author’s life plus a fixed term, which is why old paintings, historical photographs, and archive material end up here with nothing left to clear.
These sources are slower to search and the files are less polished than a stock photo. For a post about history, science, or anything with an archive behind it, they beat a generic photo of someone smiling at a laptop.
What actually causes problems
You can follow the license and still have an issue, because other rights sit on top of copyright.
- Faces. A photographer can release a photo for free without releasing the rights of the person in it. Be careful next to advertising or a sensitive topic.
- Logos and brands visible inside the frame. Those trademarks belong to someone else.
- Registering an image as your own mark. Nearly every free license blocks this outright.
- AI-generated images, which now flow into free libraries in volume. The training data question is not settled.
- Repost sites. Plenty of “free image collections” are scraped from elsewhere. Download from the original site.
There is an enforcement industry around this. Companies like PicRights and Copytrack scan sites automatically and send retroactive licensing demands; reported opening amounts run roughly $600 to $1,500 per image, with higher ones on record. Critics point out that those figures sit well above the actual license fee, and false positives are documented. Litigation tends to target businesses with revenue rather than hobby blogs, but the letters go out widely because settlements are the business model.
The defense is boring and it works: record where each image came from and when you downloaded it. If you can show the source page and the date, the conversation ends quickly.
We do not use stock photos here
Having written all that, a confession. This blog does not pull photos from any of these sites. Every featured image is a title card we build ourselves, because the thumbnail shows up in search results and sidebars and a reader is more likely to click a picture that tells them what the post is about than a nice photograph of a desk.
That did not put us outside licensing. Every time we picked a background for a card we had to ask whether we were allowed to use it. Screenshots meant checking that nobody else’s work was sitting inside the frame. When we wanted a recognizable logo, trademark concerns usually ended the idea. We settled on generating our own backgrounds or capturing our own admin screens.
One side lesson came free. We uploaded those cards as PNG and the home page grew to 2.8MB, almost all of it images. Converting to WebP fixed it. The same trap waits for downloaded stock photos: originals are often over 3,000 pixels wide, and uploading one untouched keeps you legal while making your site slow.
What I would do
Pick one photo site and stay there. Running across four sites means remembering four sets of terms; one set is enough. Save each download with the source in the filename. Use Lucide for icons so attribution never comes up. Use Google Fonts for type. When you need to build something, use Canva and skip anything with a crown on it.
Then resize the file before you upload it. That is the step people forget more often than the license.