Tesla v InterDigital: UKSC on FRAND forum

What did the Supreme Court actually decide?

On 27 July 2026 the UK Supreme Court gave judgment in Tesla, Inc and another v InterDigital Patent Holdings, Inc and others ([2026] UKSC 27), on appeal from [2025] EWCA Civ 193, allowing Tesla’s appeal and holding that the Delaware Court of Chancery is not an available forum for this dispute.

The constitution was Lord Sales, Lord Briggs, Lord Hamblen, Lord Burrows and Lord Kitchin, and the appeal was heard over three days, 27 to 29 April 2026. Those two holdings — appeal allowed, Delaware not available — are the load-bearing points, and they are jurisdictional. They tell you where a dispute of this shape can be litigated. They do not tell you what the answer to it is.

It is worth being precise about that distinction at the outset, because the temptation in SEP commentary is to read a jurisdiction ruling as a merits signal. An implementer that establishes an English forum has established an English forum. What it will get there, and on what terms, is a separate question that this judgment does not answer.

Why was a case about a patent pool in the Supreme Court at all?

Because the licensing vehicle in issue is not a SEP owner, and the usual assumptions about who owes a FRAND obligation to whom do not map onto it cleanly.

Avanci operates the Avanci 5G Platform, launched in August 2023. It is independent of the SEP owners, it owns no SEPs itself, and it has given no ETSI FRAND undertaking. That combination is the point. The familiar architecture of English SEP litigation runs through the undertaking a patentee gives to ETSI; a platform operator that has given none sits outside it. Whether such an operator can nonetheless be brought before the English courts, and joined to a claim for a declaration about the terms it offers, is precisely the kind of question that has to be resolved at the jurisdictional stage before anything else can happen.

The scale of the platform explains why the question mattered enough to reach the Supreme Court. Between launch in August 2023 and the first-instance hearing in July 2024, over 65 SEP owners had joined as Licensors and over 31 vehicle manufacturers had taken platform licences. This is not a marginal licensing structure; it is, for a large part of the automotive sector, the structure.

What does Avanci’s mandate actually permit it to do?

Its authority is narrow and pre-set, which is the feature that makes the jurisdictional question interesting rather than academic.

Avanci’s authority under the Master Licence Management Agreement, which is governed by New York law, is limited to offering a pre-determined Standard Patent Licensing Agreement subject to pre-approved modifications, and it cannot license a subset of the Licensors’ SEPs. In other words, the platform is an all-or-nothing offer administered by an agent with no discretion to negotiate around the edges and no power to carve out a portfolio. An implementer that wants a different shape of deal cannot get it from Avanci; it has to go to the Licensors.

And it can. Licensors may withdraw from the platform on at least six months’ prior written notice, and they retain the ability to grant bilateral licences. So the platform is neither exclusive nor permanent from the licensors’ side, while being effectively take-it-or-leave-it from the implementer’s side. That asymmetry is the commercial reality behind the litigation, and it is the part of the structure that in-house teams on both sides should be re-reading in their own agreements.

How much of the relevant portfolio is British?

Enough to matter, and not much more than that.

The Licensors were estimated to own around 170,000 SEPs declared to the 5G standard, of which about 7% — approximately 11,900 — were UK SEPs. That figure does two things at once. It confirms that UK rights are a small minority of the declared portfolio, which is the standard objection to English courts setting global terms. It also confirms that the number of UK SEPs is far from trivial in absolute terms, which is what gives an implementer something to point at when it wants an English anchor.

For practical purposes the lesson is arithmetical: before you plan around an English claim, work out whether the specific portfolios you are dealing with contain UK rights at all, and how many. A 7% UK share of a very large pool is a different proposition from a portfolio with no UK grants in it.

Why does French law keep appearing in an English SEP case?

Because the source of the FRAND obligation is a French-law instrument, and the English courts have consistently treated it as such.

The ETSI IPR Policy is governed by French law, and clauses 4.1, 6.1 and 6.1bis create a stipulation pour autrui for the benefit of third parties, as construed in Unwired Planet v Huawei [2020] UKSC 37. That is the mechanism by which an implementer, who is not a party to the ETSI arrangements, acquires rights it can enforce. It is also why the absence of an ETSI undertaking from the platform operator is such a pointed fact: the stipulation pour autrui runs from the undertaking, and Avanci has given none.

Anyone building or resisting a claim of this type should expect French-law contractual analysis to sit at the centre of it, with New York law governing the agency relationship between platform and licensors. English procedure, French substance, New York agency: that is the actual conflicts picture.

Who else thought this was worth intervening in?

A broad and not obviously aligned group, which is a fair indicator of how far the implications were thought to reach.

Written interventions were filed by the Computer & Communications Industry Association, the International Center for Law & Economics, ACT | The App Association, the Fair Standards Alliance and the Motion Picture Association. Interveners from the implementer side, the licensor-adjacent side and the content industries all considered the outcome material to them. When that happens on a jurisdiction appeal, it is usually because the participants expect the jurisdictional answer to determine bargaining positions well beyond the parties.

What does this judgment not settle?

A great deal, and the gaps are more consequential than the holdings for most licensing programmes.

It is a decision about jurisdiction and whether there is a serious issue to be tried. It is not a decision about the FRAND rate, and it is not a decision about whether the Avanci 5G Standard Patent Licensing Agreement terms are in fact FRAND. Nothing in it fixes a royalty, validates a platform rate card or condemns one.

It does not decide whether a platform operator that has given no ETSI undertaking owes any FRAND obligation at all. That is a substantive question about the reach of the French-law stipulation pour autrui and the position of a non-undertaking intermediary, and establishing that a claim may be heard in England is not the same as establishing that it succeeds.

Nor does it bind the Unified Patent Court, the German courts or the Swiss courts. A UK Supreme Court ruling on English jurisdiction and forum has no binding force in those systems. The EU’s own SEP policy work and the UPC’s developing FRAND case law proceed on their own footing, and a client with parallel UPC or German exposure should not assume this changes the calculus there.

Finally, the holding on Delaware is a holding about this dispute. It says the Court of Chancery is not an available forum here; it does not create a general rule about US alternative fora in SEP disputes.

What should you do about it now?

Treat it as a change in the menu of options rather than a change in the substantive law, and audit accordingly.

If you are an automotive, IoT or connected-device implementer negotiating a pool or platform licence, reassess whether an English declaratory claim is a realistic lever alongside — or instead of — UPC or German proceedings, and check whether any UK SEP in the relevant portfolios gives you an anchor for service. Do that portfolio check early; it is a factual exercise and it either supports the strategy or it does not.

If you are a licensor or a pool administrator, review your MLMA-equivalent agency terms, your withdrawal notice periods and your forum and governing-law clauses for exposure to English jurisdiction. Ask specifically whether limited agency authority, of the kind that cannot license a subset of a portfolio, helps or hurts you when jurisdiction is contested. And re-examine the assumption that making the platform offer available discharges your own clause 6.1 undertaking — that question is now live, and it has not been answered.

For anyone whose position turns on the detail, read [2026] UKSC 27 and the press summary in full before advising. The reportable holdings are narrow; the reasoning around service and the jurisdictional gateways is where the operational guidance sits.