Invalidity Searches for Standard-Essential Patents (SEPs): Searching a Haystack That Keeps Moving
Imagine you have to prove that a famous recipe was not actually new. The evidence is not in one tidy cookbook. It is spread across the final printed edition, several unfinished drafts, meeting notes from a committee, and hundreds of loose papers where different chefs each argued for their own version. Some of those papers are stored in a basement, and some are labelled with codes only insiders understand.
That is close to what an invalidity search for a Standard Essential Patent (SEP) feels like.
First, a quick refresher on SEPs
When you make a call on a 5G phone, stream a video, or connect to WiFi, your device follows a shared rulebook called a technical standard. These rulebooks are written by groups like 3GPP, ETSI, IEEE, ITU and IETF, where engineers from competing companies work together so that devices from different makers can talk to each other.
A patent is called "standard essential" when you cannot follow the standard without using the patented invention. That makes an SEP very valuable, because every company selling compliant products may need a licence. It also makes SEPs the subject of many disputes.
When a company is asked to pay for an SEP licence, one of the most common responses is to ask: "Is this patent even valid?" To answer that, someone has to look for prior art, meaning proof that the invention was already public before the patent's filing date. That search is where things get tricky.
Why SEP searches are not like normal patent searches
In an ordinary invalidity search, you look through earlier patents, research papers, product manuals and maybe some old websites. These sources are usually well organised, easy to date and easy to search.
With SEPs, some of the best prior art sits inside the standard making process itself. Standards bodies produce a huge amount of paper, and most of it was never written with future patent lawyers in mind. The useful material tends to hide in three places:
Final standards documents. These are the published specifications, often released in many versions and updates over the years.
Working group drafts. These are the unfinished versions of the standard, changed again and again as the group debates.
Contribution papers. These are the proposals companies submit to a meeting, saying "we think the standard should do it this way." A single meeting can produce thousands of them.
Each of these comes with its own headaches.
Challenge one: there is just so much of it
The sheer volume is hard to picture until you see it. A big standards group like 3GPP has produced an enormous number of documents over the years, easily hundreds of thousands. Every meeting has its own folder, every working group has its own agenda, and every proposal has its own number.
Then there is the format problem. Documents come as Word files, zipped bundles, old spreadsheets and PDFs. Some older files are scanned images, so a normal keyword search cannot read them at all. Websites get reorganised, links break, and old archives are moved from one server to another. Some portals limit who can see what, or have changed their access rules over time.
So the first job is often just finding out where the material lives and how it is organised. It is a bit like trying to search a library where the books are not shelved by topic but by the date of the meeting where somebody first talked about them.
Challenge two: engineers and patent drafters speak different languages
A patent claim is written in careful legal language. A contribution paper is written by an engineer who wants colleagues to approve a change to the standard. They might describe the same idea in completely different words.
A patent claim might talk about "a first node transmitting control information to a second node." The standards paper might just say "the base station sends a specific message on a specific channel," using shorthand and acronyms that only people in that group recognise.
If you search using the words from the claim, you may find nothing. The real prior art is sitting there, using different vocabulary. A good searcher has to translate the claim into how the standard talks about it, and sometimes into how an earlier generation of the standard talked about it. Someone who understands both patent claims and the technical language of the specific standard is very valuable in this work.
Challenge three: proving the date, and proving it was public
This is often the most important and most argued part of the whole exercise.
For a document to count as prior art, you generally need to show two things. First, that it was made available to the public before the patent's effective date. Second, that people interested in the field could have found it with reasonable effort. The exact legal tests differ from country to country, but the core idea is the same.
This is much harder with standards material than it sounds:
The date printed on a document is not always the date it became available. A paper may be prepared on one day, uploaded weeks later, and revised after that.
Computer file dates can be misleading. They may show when a file was copied to a new server, not when it was first published.
Many contribution papers get revised. The version in the archive today might not be the version that existed on the date that matters.
Some working group material is shared only with members. You then have to argue whether a document sent to hundreds of people from many companies, with no confidentiality duty, was really "public." Different courts and patent offices can see this differently.
So collecting the document is only half the job. You also need evidence about the document: the meeting dates, the upload records, the agenda listing, the email announcements, and sometimes statements from people who attended or who manage the archive. Internet archive snapshots can help too. Without this supporting evidence, a brilliant piece of prior art can be knocked out on the simple ground that nobody can prove when it was public.
Challenge four: the patent owner may have been in the room
Here is a twist that surprises many people. The company that owns the SEP often took part in the very meetings where the standard was written. Their own engineers may have submitted papers about the invention, and their patent attorneys may have filed the application around the same time.
That leads to a few situations worth checking:
The patent owner's own paper may be the prior art. If the company published its idea in a contribution paper and only filed a patent later, that paper might count against them. In some countries there is a short grace period for an inventor's own disclosures, and in others there is almost none. This is something to check early.
Competing papers from other companies. Someone else may have proposed the same thing at an earlier meeting, and their paper could be strong prior art.
Claims that drifted. Sometimes a company keeps a patent application alive and reshapes the claims over time so they match the standard as it evolves. If the final claims are not properly supported by the earliest filing, the effective date might move later. That can turn documents from the intervening meetings into prior art. This is a real opportunity for a careful searcher.
Challenge five: the standard is a moving target
A standard is not one document. It is a family of documents, released in numbered versions and stages, each one changing what came before. The same feature might appear in one release and be removed or changed in the next.
This matters because you need the version that existed at the right date, not the current one. A searcher must line up the patent's timeline with the standard's own timeline and then work out exactly what was written at each point. Change requests, which are formal proposals to edit the standard, are especially useful here because they show exactly when and why a feature was added or modified.
Earlier generations of a standard matter too. A feature that looks new in 5G might already exist in some form in 4G or even 3G documents. Looking backwards through earlier generations can uncover strong prior art.
The squeeze that helps the challenger
There is one bright spot for anyone challenging an SEP, and it comes from an old saying in patent law: what infringes if later, anticipates if earlier.
To claim that a product infringes an SEP, the patent owner has to say the standard practices the claim. But if the patent owner reads the claim broadly enough to cover the standard, then earlier documents describing the same technology may also fall inside that broad reading. So the more widely the owner stretches the claim to catch every product, the easier it can become to find earlier material that fits it as well. The patent owner has to be careful to walk a narrow path, and a good searcher will read the owner's own arguments closely for exactly this reason.
Practical tips for a better search
Every case is different, but these habits tend to pay off:
Build a timeline first. Note the patent's filing and priority dates, the relevant standards meetings, and the release dates. Everything else hangs on this.
Understand the claim in plain words. Write out what the invention does in simple language, then think about how a standards engineer would describe the same thing.
Use the standards body's own records. Declaration databases, meeting reports, agendas and change request histories often point you straight to the right meeting and working group.
Look at more than the final standard. Drafts and contribution papers are often where the earliest description lives.
Collect evidence about dates while you search. Save the document, the folder it sat in, the meeting details and any archive records at the same time. Trying to prove a date months later is far harder.
Check for different versions. Make sure the version you rely on is the one that actually existed before the critical date.
Do not stop at standards material. Academic papers, earlier patents and real products may be just as strong and easier to date.
Bring in the right people. Technical experts who attended meetings, or who know the standard well, can save enormous time and help explain the material.
Why this matters beyond one dispute
SEP disputes now play out in courts and patent offices all over the world, including Europe, the United States, China and India. In each place the rules on what counts as prior art and how it must be proven differ a little. A strategy that works in one country may need adjusting in another.
There is a larger point too. Standards make modern technology possible, and SEPs sit at the centre of how those standards are paid for. If weak patents are allowed to stand simply because the prior art is too hard to find, everyone pays more than they should. A thorough and honest invalidity search helps keep the system fair, for patent owners with strong patents and for companies who should not be paying for weak ones.
Final thoughts
Searching for SEP prior art is part detective work, part translation and part record keeping. The evidence is out there, but it is scattered, written in unfamiliar language and hard to date. The people who do it well are patient, curious and organised, and they treat every document as both a clue and a piece of evidence.
If you work in this space, we would love to hear how you approach it. What has been your toughest source of prior art to find or prove?
Comments