How to prove you published it first

The evidence, strongest first

A copyright registration

The strongest, and the only route to statutory damages in a US court. Timing matters: registration before the infringement began, or within three months of first publication, affects what you can recover. It is not required to send a notice.

An independent public archive

A Wayback Machine capture of your page from before theirs existed. Free, third-party, timestamped, and you did not create it — which is exactly why it is persuasive.

Original files with embedded metadata

For photographs especially. Camera-written capture date, device, lens and settings. Very hard to fabricate, and the thief has none of it.

Version control or CMS history

Git commits, post revision history, scheduled publish records. Weaker because you control the system, but useful supporting detail.

Dated third-party references

A newsletter that went out, a social post linking the page, an email to a client. Each one puts a date on the content from outside your own systems.

Build the record before you need it

All of this is far cheaper to do in advance than to reconstruct under pressure.

• ✓Submit your important pages to the Wayback Machine when you publish them. It takes seconds and it is free.
• ✓Keep original photography files, not just the web-sized exports. Back them up with metadata intact.
• ✓Get written assignments from every freelancer and photographer. One paragraph, signed, at the time you hire them.
• ✓If a piece of content is commercially central — a flagship guide, a catalogue of original photography — consider registering it.
• ✓Keep a simple publication log: URL, date, author. Boring, and worth a great deal the one time you need it.

What to do when they claim it is theirs

Do not argue with them. Give the reviewer the comparison and let the dates do the work. One well-built paragraph beats three pages of grievance.

A useful shape: “Our page at [URL] was published on [date] and is archived at [archive link] from [date]. The page at [their URL] reproduces [n] passages from it verbatim, and the earliest archive capture of their page is [date].”

If the dates genuinely are not on your side, stop. A notice filed against the actual author, by someone who has realised mid-way that they might be wrong, is precisely the knowing misrepresentation § 512(f) exists for.

How a reviewer actually decides

It helps to picture who reads this. Not a judge, and not a copyright specialist — usually someone at a hosting company working through a queue, with a few minutes per complaint and no appetite for adjudicating a dispute between two strangers.

That person is looking for one thing: can I verify this quickly enough to act without risk? An archive link they can click, a date, and two passages that plainly match gets a yes. A long narrative about a former business partner does not, however true it is.

So the format matters as much as the evidence. Lead with the comparison. Put the dates in the first two lines. Link the archive capture rather than attaching a screenshot. Leave the history out entirely.

The common ways priority arguments collapse

• ✓The site was rebuilt. A migration can wipe publication dates and leave every page looking new. Archive captures survive this; your CMS may not.
• ✓The content was syndicated. If you published the same text on a partner site, a marketplace and your own domain, work out which came first before claiming any of them.
• ✓A freelancer reused it. Writers sometimes sell near-identical copy to several clients. The other site may have bought it legitimately, and your claim is then with the writer rather than with them.
• ✓Both copied a third source. Supplier descriptions and manufacturer photography sit under many sellers at once. Neither of you is the author.
• ✓The date is yours alone. Evidence you control is weaker than evidence you do not. One independent capture outweighs a folder of your own screenshots.

Each of these is better discovered by you, now, than by the other side in a counter-notice. Our ownership record exists for exactly this reason: writing down how you know it is yours forces the question while it is still cheap to answer.

A worked example

Say you run an ecommerce store and discover a competitor using four of your product descriptions. You wrote them in 2024 and you have no registration, no archive captures and no particular records. This is the ordinary situation, not the unlucky one.

1. 1

   Check the archive anyway

   Search the Wayback Machine for your own URLs. Sites get crawled without anyone requesting it, and there is a reasonable chance captures already exist from before their store launched. This is free and takes two minutes.

2. 2

   Date their side as well

   The earliest capture of their page, and their domain registration date from a public RDAP lookup, both put a floor under when they could have published. A domain registered last year cannot have hosted the original of something you published in 2024.

3. 3

   Pull what your systems do have

   Product creation dates in your store admin, order history showing the listing was live and selling, the invoice from the copywriter if there was one.

4. 4

   Write two sentences, not two pages

   “Our page has been live since [date], archived at [link]. Their domain was registered [date] and its earliest capture is [date].” A reviewer can check both links and be done.

That is usually enough. Priority arguments feel like they need a lawyer, and most of the time what they actually need is two public links and a date.