EPO Enlarged Board of Appeal Brings On-Sale Bar Doctrine Closer to United States’

In its decision of July 2, 2025, the Enlarged Board of Appeal (EBA) of the European Patent Office (EPO) in G 1/23 introduced a shift in European patent law’s treatment of publicly available products, especially where their composition is not fully disclosed or reproducible. G 1/23 has broad implications for patent practitioners across all technical fields, as it reshapes the boundaries of what constitutes prior art under the European Patent Convention and affects how publicly available products, product descriptions, and marketing materials, are assessed in determining novelty and inventive step.

While this decision brings Europe closer to the United States’ on-sale bar doctrine, important distinctions remain. This update summarizes the background of G 1/23, the key findings of the decision, highlights important differences between the European and U.S. on-sale bar, and outlines practical takeaways for patent practitioners.

G 1/23 Reinterprets the Prior Art Standard

Following the previous EBA G 1/92 decision, the prevailing understanding was that for a product made available to the public (for example, a product sold on the market) to be considered part of the prior art, it had to be both analyzable and reproducible by the skilled person. In practical terms, this meant that if a product’s internal composition or structure was too complex or obscured to be reverse engineered without undue burden, it might not qualify as prior art, even if it was commercially available.

G 1/23 overrules this interpretation. The EBA held that once a product is available for public inspection or use, it becomes part of the state of the art, and any technical information that is derivable from it, such as its measurable or observable properties, must also be considered in novelty and inventive step assessments.

The patent at issue (EP 2 626 911) concerns an encapsulating material for solar cell modules. Claim 1 defines a polymeric composition having the following features: (1) Ethylene content of 80–90 mol% and C3–C20 ?-olefin content of 10–20 mol%, (2) Melt flow rate (190 °C/2.16 kg, ASTM D1238): 10–50 g/10 min, (3) Density (ASTM D1505): 0.865–0.884 g/cm^3, (4) Shore A hardness: 60–85, (6) Residual aluminum content: 10–500 ppm.

ENGAGE® 8400, a Dow (third party) manufactured ethylene/?-olefin copolymer, was cited as the closest prior art. This polymer was available on the market before the filing date of the patent and had technical properties overlapping those of the claimed invention, except for one feature (6): it had a residual aluminium (Al) content of approximately 4.4 ppm, which is below the claimed minimum of 10 ppm.

While ENGAGE 8400’s basic properties and intended applications were publicly documented, its exact composition and manufacturing process were not disclosed and could not be straightforwardly reproduced. The patentee argued that because the skilled person could not reproduce the product’s internal structure, it should be excluded from the prior art under Article 54(2) EPC. The Enlarged Board rejected that view. It reasoned that excluding such products would lead to an untenable legal fiction that is, treating a well-known, widely used product as if it had never existed from the perspective of the skilled person.

Thus, public accessibility alone is determinative. It no longer matters whether the skilled person could have discovered how to make the product. Rather, if the product was made available by being sold, distributed, or otherwise accessible without confidentiality restrictions it is considered part of the state of the art. Additionally, any information about that product that was made publicly available, such as marketing brochures, datasheets, or technical performance specifications, also constitutes prior art even if the full recipe or internal structure of the product is not disclosed or cannot be independently derived.

In sum, G 1/23 reorients the analysis under Article 54(2) EPC toward what the public has access to, not what the skilled person could technically reproduce. This has important implications for risk assessments in freedom-to-operate and invalidity strategies before the EPO.

U.S. On Sale Bar

In the United States, 35 U.S.C. §?102 bars the grant of a patent if the claimed invention was “on sale” before the critical date, which is defined as one year prior to the effective filing date of the patent application. This means that if the invention was the subject of a commercial offer for sale more than one year before the application was filed, the patent may be invalid. In Pfaff v. Wells Electronics, Inc., 525 U.S. 55, 65 (1998), the Supreme Court clarified that an invention is on sale if it is the subject of a commercial offer for sale and is ready for patenting. It also clarified that the invention is ready for patenting when a manufacturer has already reduced it to practice; a composition of matter is reduced to practice when it is completely composed. Furthermore, the statutory on-sale bar is not subject to exceptions for sales made by third parties either “innocently or fraudulently”. The fact that these sales were not made by the patentee is therefore irrelevant. See Abbott Laboratories v. Geneva Pharmaceuticals, Inc., 182 F.3d 1315 (1999). Later, in Helsinn Healthcare S.A. v. Teva Pharmaceuticals USA, Inc., No. 17-1229 (2019), the Court held that even secret or confidential sales can trigger the on-sale bar, so long as the sale itself places the invention “on sale” in the commercial sense.

Europe v. U.S. On Sale Bar

There are important differences between Europe and the United States in the scope and application of the on-sale bar. Under U.S. law, even confidential or secret sales may trigger the bar, whereas under the EPC, only public availability counts. Moreover, Europe does not offer a general grace period except for limited exceptions under Article 55(1)(a) EPC like disclosures due to abuse or at official exhibitions while the United States provides a one-year grace period under 35 U.S.C. §102(b). Experimental use exceptions also differ. In the United States, experimental use may avoid the on-sale bar if the purpose is to test the invention’s functionality. In Europe, experimental use may avoid creating prior art only if it occurs under conditions of confidentiality or control (see for example T 152/03 and T 229/86). These differences can lead to divergent outcomes for similar fact patterns and must be considered when developing global patent strategies.

Implications Across All Technical Fields

Although G 1/23 arose from a case in polymer chemistry, its impact extends to all fields of technology. The Enlarged Board itself noted that the question of reproducible prior use can arise in any area involving complex products. By effectively broadening what counts as prior art, G 1/23 will influence how novelty and inventive step are evaluated across the board. Patent practitioners should be aware of several key implications:

No ‘Safe Harbor’ for Secret Know-How Once Product is Public: Pre-G 1/23, some applicants might have assumed that if they kept their manufacturing process or composition details secret, they could market a product and later seek a patent on it (or on its properties) so long as competitors couldn’t readily reproduce it. This strategy is now untenable. Any non-confidential sale or public use of a product before the filing date will count against novelty or inventive step of that same product in Europe. (Unlike in the United States, purely confidential sales still may not create prior art in Europe, the product may need to enter the public domain).

Public Disclosures of Product Information are Fully Prior Art: Companies often publish data sheets, white papers, or brochures describing features of their products. G 1/23 makes clear that any technical information disseminated about a product is part of the state of the art. Even if certain internal aspects are not disclosed, whatever is disclosed (or can be measured from the product) will be considered known to the public. For practitioners, this means that searching prior art should include not only patents and literature, but also product brochures, datasheets, and other commercial publications for relevant features.

Relevance in High-Tech and ‘Black Box’ Scenarios: The decision is especially significant for fields like pharmaceuticals and biotechnology, where the exact properties or structure of a drug might be hard to ascertain from the marketed product. Similarly, in fields like electronics or software, a device sold to the public might be a prior art for what it does or contains, even if its source code or schematic is not publicly known. A patent on an algorithm secretly embedded in a marketed device, for instance, could be challenged if the device’s behaviour inherently disclosed the algorithm’s output.

Focus on Evidence: The Enlarged Board’s approach shifts the debate to proof of properties. If a prior product potentially anticipates or renders an invention obvious, the discussion will turn on whether the opponent has adequate evidence of the product’s relevant features. For example, in ENGAGE 8400, the aluminium content was measured and could show it was outside the claimed range (4.4 ppm). Under G 1/23, if a property was not publicly documented and cannot be reliably measured or inferred, the patentee may still prevail for lack of proof; but this will be argued as an evidentiary gap.

Inventive Step Analysis: While G 1/23 closes the door on patenting an identical product that was previously available, it does not preclude patents on improvements or modifications of such products. For example, as in G 1/23, while ENGAGE® 8400 was the closest prior art, the claimed higher aluminum content was not in the product. Now that G?1/23 clarified that ENGAGE® 8400 is prior art regardless of its reproducibility, the case will return to the Technical Board of Appeal (T?0438/19) for a final decision. The key question for the Board will be “Would a skilled person, starting from ENGAGE® 8400 and the available knowledge about it, have found it obvious to adjust the aluminium content to the claimed range, thereby achieving the improved properties?”. Thus, practitioners need to focus on the technical effect of the distinguishing feature both when drafting the patent application and when opposing or defending the patent.

Leverage Prior Uses in Oppositions/Litigation: If you are opposing a patent or defending against one, consider whether any prior product sales or uses can serve as invalidating prior art. Even if the product’s exact composition wasn’t known, any aspect of it that was publicly available (by demonstration, analysis, or documentation) can be used in argument. Be prepared to provide evidence such as lab analyses, affidavits, dated product literature to establish the features of the prior product. The burden is on the opponent to prove relevant properties. It may be prudent to review any granted European patents that were predicated on a rejected interpretation of G 1/92. Patents granted in the past because a prior use was deemed non-enabling might now be more vulnerable if challenged.