Copyright and Google Images: What UK Businesses Need to Know Before Using a Photo

Images found through Google Images are not free to use. Google Images is a search index of pictures hosted on other people's websites, not a stock library. Almost every photograph on the internet is protected by copyright automatically, and in the UK that protection lasts for the life of the photographer plus 70 years.

Using an image on your website, in an advert or on social media without a licence is copyright infringement, even if you found it through Google, even if you credited the source, and even if you took it down as soon as you were challenged.

The only images you can safely reuse are ones you own, ones you have licensed, ones released under a licence whose terms you have actually read, or ones genuinely in the public domain.

Why so many businesses get this wrong

Google Images looks like a library. You type a word, you get a grid of pictures, you right click and save. Nothing in that experience feels like theft, and that is precisely the problem.

What is actually happening is that Google is showing you thumbnails of images that live on other people's servers. Google does not own them, does not license them to you, and offers no indemnity if the owner comes after you. The interface has been quietly nudging users towards this understanding for years. In February 2018, Google removed the "View image" button from image results as part of a settlement with Getty Images, which had complained that the button made it too easy to grab a high resolution file without ever visiting the site that hosted it. Google also made the copyright warning beneath each preview more prominent at the same time.

That warning is still there. Most people scroll straight past it.

Definition: copyright in a photograph. In the UK, a photograph is an "artistic work" under the Copyright, Designs and Patents Act 1988. Copyright arises the instant the shutter is pressed. There is no registration, no fee, no copyright symbol required, and no public register you can check. If a photograph exists, assume someone owns it.

What UK copyright law actually says about images

UK copyright is governed by the Copyright, Designs and Patents Act 1988. Four points matter most for anyone putting pictures on a website.

Protection is automatic. The Intellectual Property Office is explicit that copyright is an automatic right with no official formalities and no government register in the UK. You do not need to see a copyright notice for copyright to exist.

Protection lasts a very long time. For artistic works including photographs, copyright expires 70 years from the end of the calendar year in which the author dies. A photograph taken in 1990 by someone who is still alive will remain in copyright well into the next century.

Ownership is not always obvious. The first owner is normally the photographer, but where a photograph is taken by an employee in the course of employment, the employer owns it. Commissioning a photographer does not automatically transfer copyright to you unless the contract says so.

Moral rights sit alongside economic rights. The author has the right to be identified and to object to derogatory treatment of the work. Cropping a photographer's watermark out is not a neutral act.

The UK has a narrow set of exceptions rather than the broad American doctrine of fair use. Fair dealing exceptions cover things like noncommercial research and private study, criticism and review, quotation, and reporting current events, and each carries its own conditions including sufficient acknowledgement. Illustrating a blog post about your product range with someone else's photograph does not fall into any of them.

The Usage rights filter: what it does and what it does not do

Google does provide a tool for finding reusable images, and it is genuinely useful as a starting point. Under Tools in Google Images there is a Usage rights dropdown with three options.

Alongside the filter, eligible results carry a Licensable badge in the corner of the thumbnail. Opening the image shows a "Licence details" link and, where the publisher has provided one, a link to a page where the image can be bought. The badge is driven entirely by metadata: either schema.org structured data on the hosting page or IPTC photo metadata embedded in the file itself.

Here is the crucial limitation. Google filters by licence information supplied by the sites hosting the images, and its own help page tells users to confirm that any licence is legitimate and to check the exact terms before reusing anything. Google explicitly advises users to always confirm an image's licence information.

In other words, the filter tells you what a website has claimed. It does not tell you whether that claim is true. A site with no authority whatsoever to license an image can still tag it as Creative Commons, and Google will surface it in the filtered results. Nor does the filter flag attribution obligations, model releases, property releases or trade mark issues in the frame.

Treat the Usage rights filter as a shortlist, not a clearance.

Creative Commons is a licence, not a free for all

A great deal of trouble comes from people treating "Creative Commons" as a synonym for "free". It is not. It is a family of six licences plus a public domain dedication tool, and every one of the six requires attribution to the creator.

Two traps catch UK businesses repeatedly. The first is the NonCommercial element: a photograph on a company website is a commercial use even if the specific page does not sell anything. The second is attribution: dropping a CC BY image into a blog post without naming the photographer and linking the licence is a breach of the licence, which means the licence falls away and you are back to plain infringement.

Creative Commons licences also cannot be revoked once granted, which cuts the other way and is worth knowing if you are considering releasing your own work.

Copying versus embedding: does hotlinking help?

Some people assume that embedding an image from its original location, rather than downloading and rehosting it, avoids the problem. The legal position is more nuanced than that assumption allows, and it is not a strategy to rely on.

European case law, which continues to influence UK courts on retained principles, has drawn a line around the concept of a "new public". In Svensson and BestWater the Court of Justice held that linking to or embedding content that the rights holder had already made freely available did not reach a new public and so was not a fresh communication to the public. But in Land Nordrhein Westfalen v Renckhoff, the court held that downloading a freely available photograph and reposting it on a different website did reach a new public, because the audience the rights holder had in mind was the audience of the original site.

The practical takeaway for a UK business is simple. Saving an image and uploading it to your own media library is the highest risk act you can take. Embedding sits in more contested territory that depends on how the content was made available and whether any technical restrictions were circumvented. Neither route gives you a licence, and neither protects you from a claim if the image was posted without the owner's permission in the first place.

What it costs when it goes wrong

Civil claims and damages

The usual starting point for damages is the notional licence fee: what you would have paid had you asked. That figure alone is rarely the end of it.

In Absolute Lofts South West London Ltd v Artisan Home Improvements Ltd a Bradford building firm copied 21 photographs of completed loft conversions from a London competitor's website and put them on its own. The notional licence fee was assessed at £300, reflecting what equivalent stock images would have cost. The court then awarded a further £6,000 in additional damages for flagrancy and unfair profits, because the photographs had misrepresented to customers that Artisan had carried out work it had not done. Twenty times the licence fee, on an admitted claim, over pictures that were not even professionally shot.

Most image claims never reach a courtroom. Those that do often land in the Intellectual Property Enterprise Court small claims track, which handles the simplest intellectual property disputes worth less than £10,000, runs short and informal hearings, and rarely orders the loser to pay the winner's costs. It exists precisely because photographers needed a proportionate route to enforcement.

Demand letters from monitoring agencies

Far more common than litigation is the letter. Services such as Pixsy, Copytrack and ImageRights run reverse image search across the web on behalf of photographers and issue retroactive licensing demands when they find a match. Pixsy alone reports having handled more than 200,000 cases of unauthorised use, and typical demands reported by UK recipients sit in the several hundred pounds per image range.

These letters are not automatically valid, and they are not automatically bogus either. The government publishes guidance on letters alleging online copyright infringement that is worth reading before you respond to one. What you should not do is ignore it, admit liability, or pay immediately without checking whether the sender can actually establish ownership and whether the sum demanded bears any relation to a real licence fee.

Criminal liability

Criminal sanctions exist but are aimed at commercial scale piracy rather than a misjudged blog post. Since 1 October 2017, section 107 of the Copyright, Designs and Patents Act 1988 has carried a maximum custodial sentence of ten years for online copyright infringement, raised from two years by the Digital Economy Act 2017. The offence requires an intention to make a gain, or knowledge that the act will cause or risk loss to the rights owner.

Where to get images you can actually use

Unsplash was acquired by Getty Images in 2021, though the free tier licence terms were not materially changed. The wider point about all free stock platforms holds: they generally offer no indemnification and do not verify that uploaders own what they upload, which means an image can be taken down and a claim can still follow.

AI generated images: where the law stands in 2026

Two developments have shaped the position for UK businesses.

In November 2025 the High Court handed down judgment in Getty Images (US) Inc & Ors v Stability AI Limited [2025] EWHC 2863 (Ch). Getty abandoned its primary copyright claims during trial, and the court rejected the remaining secondary copyright infringement claim, finding only historic and extremely limited trade mark infringement. The judgment gave very little clarity on whether training AI models on copyright works is lawful, because the issues before the court had narrowed so far.

Then in March 2026 the government published its statutory Report on Copyright and Artificial Intelligence under sections 135 and 136 of the Data (Use and Access) Act 2025. The outcome was to maintain the status quo: no new legislation, no new regulator, and existing law applied by the courts. The originally preferred option of a text and data mining exception with a rights holder opt out was dropped after consultation responses ran heavily against it, with 88% of respondents supporting a requirement for licences in all cases.

For a business generating images with AI, the practical risks are unchanged and worth stating plainly. Output that closely resembles an identifiable existing work can still infringe. Output containing recognisable watermarks or trade marks creates a separate exposure. Ownership of purely machine generated output is legally uncertain, which matters if you need to stop a competitor reusing your visuals. Check whether your provider offers indemnification, because most free tools do not.

On the detection side, Google now reads Coalition for Content Provenance and Authenticity metadata to power its About this image feature across Search, Images and Lens, surfacing whether an image was created or edited with AI tools where the provenance data is present. Coverage is still patchy because the metadata has to be there in the first place, but provenance signals are becoming part of how images are assessed.

A practical clearance workflow

Adopt this before anything goes live and most of your exposure disappears.

1. Default to owned or paid. Your own photography and paid stock cover the overwhelming majority of business needs and remove the argument entirely.
2. Never save directly from Google Images. Use the Usage rights filter to find candidates, then click through to the hosting site and verify the licence there.
3. Read the actual licence, not the label. Confirm commercial use is permitted, confirm whether edits are allowed, and confirm the attribution wording required.
4. Record the evidence. For every image, keep the source URL, the licence type, the date acquired, the invoice or licence number, and a screenshot of the licence terms. If a demand letter arrives in three years, this file is your defence.
5. Attribute properly where required. Name the creator, name the licence, and link to both.
6. Audit what is already published. Run a reverse image search on the images currently on your site, or check them against your records. Legacy pages and old blog posts are where unlicensed images hide.
7. Brief anyone who touches the website. Agencies, freelancers and junior marketers are the usual entry point. Put the rule in writing and make it a contractual obligation for suppliers.

Protecting your own images from the other side

If your business produces original photography, the same tooling works in your favour.

Add image licence metadata to your site so your images qualify for the Licensable badge in Google Images. This means adding an ImageObject structured data block with the license property, and ideally acquireLicensePage, creator, creditText and copyrightNotice. Alternatively, embed IPTC photo metadata directly in the file, using the Web Statement of Rights field, which travels with the image from page to page. Where both exist and conflict, Google uses the structured data. Participation is not a ranking factor, but it does put a licensing route in front of anyone who finds your work.

Beyond that: keep original raw files with intact EXIF data as proof of authorship, run periodic reverse image searches on your key assets, and issue a clear, proportionate letter before action if you find unauthorised use. The IPEC small claims track exists for exactly this.

Frequently asked questions

Can I use Google Images for free? No. Google Images is a search index of pictures hosted on other websites, not a stock library. Google does not own the images, does not grant you a licence, and provides no indemnity. You need permission from whoever owns the copyright in each image.

Is it legal to use a Google image if I credit the source? No. Attribution is not a substitute for permission under UK law. Crediting a photographer you never asked does not create a licence. Attribution only matters when a licence, such as a Creative Commons licence, specifically requires it as one of its conditions.

How do I find images on Google I am allowed to use? Open Google Images, select Tools, then the Usage rights dropdown, and choose either Creative Commons licences or Commercial and other licences. Then click through to the hosting website and verify the licence terms yourself, because Google reports what the host site has claimed rather than verifying it.

What does the Licensable badge on Google Images mean? It means the publisher has attached licensing metadata to the image, either as schema.org structured data on the page or IPTC metadata in the file. Selecting the image shows a Licence details link and often a link to buy a licence. The badge indicates a licence exists, not that you already hold one.

How long does copyright last on a photograph in the UK? For artistic works including photographs, copyright expires 70 years from the end of the calendar year in which the photographer dies, under section 12 of the Copyright, Designs and Patents Act 1988. Practically, almost every photograph you find online is still in copyright.

What happens if I use a copyrighted image by accident? You are still liable. Copyright infringement does not require intent. In practice you will usually receive a demand for a retroactive licence fee. Courts can award the notional licence fee plus additional damages where the infringement was flagrant, as in Absolute Lofts v Artisan Home Improvements, where £300 in licence fees was followed by £6,000 in additional damages.

Does removing the image solve the problem? Not by itself. Taking the image down stops the infringement continuing but does not extinguish the claim for the period it was published. Remove it, preserve your records of where you obtained it, and take advice before responding to any demand.

Can I use AI generated images commercially in the UK? You can, but with caveats. Output resembling an identifiable existing work can still infringe, ownership of purely machine generated output is legally uncertain, and most free AI tools offer no indemnity. The government's March 2026 report on copyright and AI kept existing law unchanged rather than creating new certainty.

Are images from Unsplash, Pexels and Pixabay safe for business use? Their licences permit commercial use without attribution, which covers most needs. The risk is that these platforms do not verify that uploaders own what they upload and offer no indemnification, so a claim can still follow if an image was uploaded by someone who had no right to share it. Keep a dated record of every download.

What should I do if I get a copyright demand letter? Do not ignore it, do not admit liability, and do not pay immediately. Check whether the sender can establish ownership and standing, check whether the sum bears any relation to a genuine licence fee, gather your own records of where the image came from, and read the government's guidance on letters alleging online copyright infringement before responding.

The short version

Google Images is a finding tool, not a licensing tool. The filters help you narrow the field, and the Licensable badge points you towards a legitimate purchase, but nothing in Google's interface transfers a right to you. UK copyright is automatic, long lived and enforced by an increasingly automated detection industry.

The cost of getting it right is a stock subscription and a spreadsheet. The cost of getting it wrong runs to thousands of pounds per image, arrives without warning, and lands on work you published years ago.

This article is general information about UK copyright law and does not constitute legal advice. Seek advice from a qualified intellectual property solicitor on your specific circumstances.