Future-proofing your licenses: Key changes in the new EU technology transfer regime

The TTBER itself is left almost unchanged. The 20% threshold for competitors and the 30% threshold for non-competitors (Article 3(1) and 3(2) TTBER) – the combined market-share ceilings below which block-exemption protection applies – survive untouched. The same goes for the excluded restrictions under Article 5 TTBER, which still remove exclusive grant-back obligations (Article 5(1)(a)) and no-challenge clauses (Article 5(1)(b)) from the benefit of the block exemption. And the catalogue of "hardcore restrictions" under Article 4 TTBER, which strips block-exemption protection where an agreement contains a core restraint such as price fixing (Article 4(1)(a) for competitors; Article 4(2)(a) for non-competitors), output limitation (Article 4(1)(b)), or allocation of markets or customers (Article 4(1)(c) and 4(2)(b)), also remains intact – albeit with three tidying edits:

1. spare parts are protected only where the licensee incorporates the contract product into its own product (Article 4(1)(c)(iii) TTBER), narrowing a carve-out that used to cover stand-alone spare-parts trade;
2. the selective-distribution carve-out for passive sales restrictions (Article 4(2)(b)(iii) TTBER) now attaches to territory rather than membership, mirroring the language of the Vertical Block Exemption Regulation; and
3. each hardcore limb now expressly refers to "the contract products”, closing the drafting ambiguity as to whether hardcore treatment reaches goods not covered by the licensed technology.

Two changes are relatively minor but matter commercially.

1. The grace period for a threshold breach (Article 8(e) TTBER) – the window during which an agreement remains block-exempt even after the parties' combined market share crosses the threshold – moves from two to three consecutive calendar years, harmonising the TTBER with the Vertical Block Exemption Regulation (Reg. (EU) 2022/720) ("VBER"). In dynamic technology markets – think AI, next-generation battery chemistry, quantum components – an extra twelve months of block-exemption runway is real comfort.
2. Furthermore, there is now a zero-market-share rule for pre-commercial technologies (Recital 13, Article 8(d) TTBER), meaning that a technology that has not yet generated sales of contract products is explicitly treated as holding a zero market share. Because the market-share test is normally run on sales of products incorporating the licensed technology (Article 8(a) TTBER), a technology still in pre-commercial development produces no data points at all – and the 2014 regime reached the same outcome only through a single paragraph in the old Guidelines. The absence of a Regulation-level anchor created awkwardness in pharmaceutical and life-sciences licensing, where deals often concern nascent technologies waiting for regulatory "go-to-market" approval. That has now been dealt with.

Everything else the Commission has revised and added during the reform process, however, sits in the Guidelines. Let's take a look.

The 2014 Guidelines did not mention data. Not by accident: data licensing had grown into a commercial category between drafting cycles, and the 2014 framework simply had no place for it. The new Guidelines correct that with a three-layer framework.

Layer one: the TTBER applies directly to data licensing where the licensed data qualifies as know-how under Article 1(1)(i), as one of the listed technology rights under Article 1(1)(b), or as an ancillary element of a technology transfer agreement that meets Article 2(3).

Layer two: where data is licensed outside the TTBER but as part of a database protected by copyright or the sui generis right under Directive 96/9/EC, the Commission applies the TTBER and Guidelines principles "by analogy."

Layer three: for any other data, case-by-case assessment – with explicit acknowledgment that the guidance may not apply in all cases, since data licensing practice is evolving faster than the Commission can codify.

The information-exchange dimension receives careful attention. Where a data licence itself falls outside Article 101(1) TFEU, information exchange ancillary to the licence is safe – provided the exchange is objectively necessary and proportionate to the licence's purpose. Where that ancillary test fails, or where the information exchange is itself the agreement's main object, Chapter 6 of the Horizontal Guidelines applies, with HSBC Holdings (C-883/19 P) cited as authority for the by-object standard.

LNGs – arrangements under which technology implementers jointly negotiate the terms of technology licences they wish to obtain from technology holders – received no Guidelines treatment in 2014. They do now.

The decisive (and, in practice, contentious) structural question is how to distinguish a genuine LNG from a buyer cartel. The Guidelines answer this by describing how an LNG can be permissible when it interacts with the technology holder collectively but on behalf of identified members, defines its form, scope, and functioning in a written agreement, and confines itself to the negotiation of licence terms rather than coordinating members' downstream market conduct or exchanging commercially sensitive information on their individual market behaviour. Genuine LNGs that meet those criteria do not, in the Commission's view, restrict competition by object, leaving room for an effect-based assessment under Article 101(1) TFEU.

That effects assessment runs on a 15% / 15% dual threshold – combined demand share on the relevant technology market, combined supply share on the relevant product market. Below both, market power is unlikely and the LNG generally falls outside Article 101(1) TFEU. Above either, the full effects analysis follows, with particular attention to coordinated delays in licensing negotiations: the Commission will treat those as an aggravating factor within the overall LNG assessment rather than as a standalone violation.

For LNGs above the thresholds – and for those who want to mitigate their competition law risks further – the Commission sets out a list of measures that make an Article 101 TFEU challenge less likely. That list reads like a draft LNG membership charter: open and non-discriminatory membership, transparent disclosure of rules and the negotiation role, activity limited to joint negotiation, restrained information exchange, technology holders remaining free to negotiate with third parties in parallel, any standstill on bilateral member-by-member negotiations capped at six months, and licensing fees sought not exceeding 10% of the product sale price.

The Commission is not starting from zero in that regard. In July last year it issued an informal guidance letter on the Automotive Licensing Negotiation Group ("ALNG"), signalling that joint negotiations over SEP licences could, under the right conditions, comply with EU antitrust rules. The conditions the Commission set then – open membership, voluntary engagement for SEP holders, information exchange limited to what the negotiation requires, and a 15% cap on combined demand share in the upstream licensing market (see our article here) – are now the spine of the new Guidelines framework.

Notably, the draft Guidelines of September 2025 had offered a formal safe harbour for LNGs that met those conditions. The final text, however, does not. In a separate explanatory note, the Commission states why: enforcement experience is too thin, and the conditions risked being either under-inclusive (missing real concerns) or over-inclusive (potentially deterring pro-competitive LNGs) – a statement that is both true but also glossing over the fact that the draft safe harbour had faced heavy criticism above all from companies active in the licensing of patents. The substance of the draft safe harbour therefore survives – but as measures LNGs can (and should) take proactively to avoid infringing Article 101(1) TFEU, not as a definitive “protected zone”.

Whether that distinction will actually matter in European enforcement practice may be questioned. What is certain, however, is that the US is taking a stricter stance on the whole topic, with the Department of Justice having opened an investigation in March 2026 into the ALNG, on the theory that such coordination could amount to a buyer cartel under US law. Which means that an LNG structured to fit neatly within the new EU Guidelines may still face antitrust exposure across the Atlantic – a divergence that clients operating globally will need to price in from day one.

The Guidelines also address two additional areas of enforcement refinement.

For technology pools, the 2014 Guidelines listed seven conditions for a pool to fall presumptively outside Article 101(1) TFEU. The new Guidelines now list eight. The additional condition is a disclosure obligation: pool operators must disclose the rights included in the pool in an effective manner – at minimum the patent or application number and country of registration, with updates on request. Furthermore, the already existing essentiality condition is expanded to require that the methodology used to assess essentiality is itself disclosed (scope, content, sampling technique, assessor-selection criteria). And the FRAND licensing condition is expanded to prohibit "double dipping" – charging licensees twice for the same technology rights, once bilaterally and once via the pool.

Beyond that, the new Guidelines add a dynamic essentiality principle that has no direct equivalent in the 2014 text. Technologies that become non-essential over time – through standard evolution or the emergence of third-party alternatives – must either be removed from the pool or offered in a reduced-royalty licence without the now-non-essential component. Technology pools can no longer treat essentiality assessment as a one-time exercise at pool creation.

While the 2014 Guidelines devoted a mere two sentences to pay-for-delay settlements, the new Guidelines give them a full sub-section and integrate the complete CJEU body of case law since 2020: Generics/GlaxoSmithKline (C-307/18), Lundbeck (C-591/16 P), Servier (C-176/19 P), Krka (C-151/19 P), and most recently Teva/Cephalon (C-2/24 P, October 2025).

The operative test derived from that judicature is now formalised in the Guidelines: A settlement between actual or potential competitors restricts competition by object where it is evident from examining the agreement that the value transfers involved have no explanation other than the parties' commercial interest not to engage in competition on the merits. Legitimate value transfers include compensation for costs or disruptions caused by the settled litigation, or remuneration for goods or services actually supplied. The Commission clarifies – following Servier and Teva – that the generic party's net gain need not exceed the profits it would have obtained by winning the patent proceedings.

This is not a surprising move from the Commission. In view of the case law and the hardened stance against pay-for-delay settlements as a phenomenon fought by competition authorities in several cases, it was to be expected that the new TTBER package would address this. It is worth pointing out, however, that pay-for-delay was not added to the TTBER's catalogue of hardcore restrictions – the Commission's way of acknowledging that whether or not a settlement actually meets the incriminating criteria set out above so that a by-object restriction is present, is a judgment that must be made on the specific facts of each agreement and its context; that analysis requires a depth that does not quite fit the catalogue of (relatively) straightforward-to-assess hardcore restrictions.

Last but not least, there are three more changes that may look technical but affect day-to-day practice.

(1) The TTBER now contains definitions of "active sales" (approaching specific customers or territories) and "passive sales" (responding to unsolicited customer requests) – lifted from the VBER. Attached to those definitions is the full online-commerce catalogue, which tells parties how to characterise online conduct: a country-code domain (.de, .fr) is typically active targeting of that country, a website available in a local language is active targeting of speakers of that language, and paid search advertising shown to users in a given territory is active targeting of that territory.

(2) The definition of "competing undertakings" has been re-anchored. The 2014 TTBER defined potential competition through a SSNIP-style test: whether a party would be likely, in response to a small and permanent increase in prices, to make the investments needed to enter the market. The new definition drops the SSNIP reference and aligns with the Commission's Horizontal Guidelines: potential competitors are those likely, within a sufficiently short timeframe, to impose competitive pressure on incumbents. The shift is subtle and the outcomes will usually align. Where they diverge, the new formulation is marginally broader – more situations may qualify as "between competitors," triggering the 20% threshold rather than the 30%.

(3) The captive-use spare-parts carve-out is narrower than before. Under the TTBER, a licensor can lawfully limit a licensee to producing the contract products solely for the licensee's own use. That restriction has always been subject to a proviso protecting the spare-parts market, in that the licensee remains free to sell spare parts actively and passively, so that downstream customers are not cut off from replacements. Under the new text, that freedom applies only where the licensee incorporates the contract product into its own product (Article 4(1)(c)(iii) TTBER). Stand-alone spare-parts trade by the licensee no longer benefits from the carve-out. Industrial automation, machinery, and after-sales-heavy sectors should review captive-use clauses against this narrower frame.

Notably absent: any explicit treatment of AI model or foundation-model licensing. The Commission's silence on that matter can be interpreted either as confidence that the current TTBER revamp will suffice to capture those as well – or that more specific regulation will come further down the road.