The importance of confidentiality (and why G 1/23 did not establish a US style on-sale bar to patentability in Europe)

G 1/23 was potentially one of the most important developments in European patent law this year. Much of the commentary relating to the potential impact of G 1/23 has made reference to a shift towards a US-style on-sale bar, with some even going so far as to proclaim that G 1/23 established something akin to a European “on-sale bar”. However, in PatKat's view, comparing G 1/23 with the US on-sale bar is unhelpful and only serves to confuse matters. When considering the practical consequences of G 1/23, one of the most important is understanding the legal requirements for protecting trade secrets with adequate confidentiality provisions (and how the EPO may assess this). Crucially, such considerations are irrelevant with respect to the US on-sale bar, for which conditions of confidentiality do not apply.

G 1/23: Case refresh

In G 1/23, the EBA was asked whether a product that was on the market should be excluded from the prior art for the sole reason that its internal structure couldn't be analysed and reproduced without undue burden at the priority date. The case underlying the referral related to a complex polymer but the decision applies to any "black-box" product that cannot be reproduced.

Mysterious prior art

The EBA found, contrary to a significant strand of previous case law, that a product that was on the market should not be excluded from the prior art because it couldn’t be analysed (IPKat). Importantly, this also means all the public technical information about that product is also considered prior art, both for novelty and inventive step, which was not previously the case.

Understanding the US "on-sale bar"

The on-sale bar is a statutory provision in U.S. patent law (35 U.S.C. § 102) that prevents an inventor from obtaining a patent if the invention was sold more than one year before the patent application was filed.

The US Supreme Court established the current legal standard for triggering the on-sale bar in Pfaff v. Wells Electronics (1998). In this case, the Court created a two-prong test. For the on-sale bar to apply, the invention must be (1) the subject of a commercial offer for sale, and (2) "ready for patenting" at the time of that offer. 

The scope of the on-sale bar received an important update in Helsinn v. Teva (2019). In this case Helsinn sued Teva for infringing its patent on a specific dose of palonosetron, a treatment for chemotherapy-induced nausea. Teva countered that the patent was invalid because Helsinn had commercially licensed the rights to the drug more than a year before filing its patent application. The Supreme Court ruled in favour of Teva, finding that even a "secret sale", where an invention is sold to a third party under a confidentiality agreement, still triggers the on-sale bar to patentability. 

G 1/23 did not establish a US-style "on-sale bar"

The US on-sale bar is thus very different to the situation in Europe, both before and after G 1/23. Importantly, "the state of the art" under Article 54(2) EPC does not encompass confidential disclosures that were not "made available to the public", including confidential commercial sales. Applying G 1/23, T 0143/24 confirmed: 

"free access on the market is lacking if the product is sold subject to confidentiality agreements. Contrary to the opponent's view, nothing to the contrary follows from decision G 1/23, as that decision refers to products that are on the market and accessible to the public. This requirement is not met if the product is marketed only under confidentiality agreements." [r. 1.3, Machine translation]

Contrary to much of the commentary on this decision, G 1/23 thus did not establish a US-style on-sale bar to patentability in Europe. Whilst the US on-sale bar applies even to sales made under conditions of confidentiality, G 1/23 very much did not introduce confidential disclosures into the prior art. Ensuring that commercial sales are made under adequate confidentiality conditions is therefore still a viable strategy in Europe.  

G 1/23 does not just apply to commercial products

Another important difference between G 1/23 and a US-style on-sale bar is that, unlike the US on-sale bar, G 1/23 relates to any product and not just to commercial sales. G 1/23 applies to any product that is publicly disclosed, regardless of whether this was as a result of a commercial sale or in some other way.

From the pharma perspective, G 1/23 therefore raises questions over the impact of disclosing pharmaceutical products during clinical trials. Of course, the general reproducibility and regulatory disclosure requirements relating to pharma products mean that it has always been generally advisable to file a patent application covering the clinical product before it enters the clinic. However, there are types of pharmaceutical inventions which are non-reproducible, do not need to be disclosed for regulatory purposes, and thus can be kept secret even when used clinically, including formulations and complex devices, as well as some advanced therapies (see e.g. T 0670/20, IPKat). Following G 1/23, if these products are not provided under conditions of confidentiality, they become full prior art with respect to any future European patent application. From this perspective, G 1/23 can therefore be considered more of an alignment with the US law on inherency, rather than the on-sale bar. 

Nonetheless, it has thankfully been confirmed that G 1/23 only introduces tangible non-reproducible products as relevant prior art, and not hypothetical "armchair" disclosures (T 1044/23). 

There is no grace period or "shield" in Europe

Another important difference between the US and Europe with respect to prior art is the existence of the 12 month grace period available to inventors in the US. There is, by contrast, no such grace period in Europe. The US grace period importantly also applies to the on-sale bar provision. Therefore, a public commercial sale of a product by the inventors within 12 months of the priority date may invalidate a patent in Europe, but not in the US. Conversely, a confidential commercial sale of a non-reproducible product 12 months before the priority date may invalidate the patent in the US, but not in Europe. 

As an aside, there is an interesting nuance with respect to the US on-sale bar on the grace period. US law includes a "shield" provision that allows an inventor's public disclosure to automatically disqualify any subsequent third-party disclosures from serving as invalidating prior art during the one-year grace period. The US Federal Circuit case in Sanho v. Kaijet (2024) considered the interaction between the on-sale bar and this protection from third party disclosures. The Federal Circuit found that whilst a private, secret sale is sufficient to trigger the on-sale bar (invalidating a patent under Helsinn), it does not count as a "public disclosure" that would trigger the safe harbour provision to protect the inventor against other prior art. As such, in the US, private commercial activities may invalidate a patent, but cannot be used as a shield to establish an early effective filing date against competitors.

Final thoughts

In terms of future outlook, G 1/23 is therefore likely to impact the decisions companies make with regards to trade secrets as a key component of IP strategy. G 1/23 reemphasises the need to have robust and secure trade secret policies, and to be clear on what information is being disclosed when marketing or developing a product and how this may affect subsequent patent filings. This is particularly important for products that might previously have been protected under the non-reproducibility case law prior to G 1/23, including software and some advanced therapies.

Additionally, as we have seen with respect to a number of prior use cases this year, having adequate confidentiality agreements in place therefore remains paramount in Europe following G 1/23. Whilst the EPO will take a nuanced approach to the evidence standard for confidentiality, an implied understanding is rarely enough (T 2027/23, T 0439/24, T 0166/24, T 1311/23, see also IPKat).  

Finally, it is essential to remember that G 1/23 is merely the new European position. A multijurisdictional IP strategy must take account of the on-sale bar in the US, which cannot be avoided with confidentiality restrictions, but also the lack of a 12-month grace period for disclosures by the inventor in Europe, as well as the applicability of G 1/23 to any disclosure not just commercial sales. If there are plans to disclose and/or sell a product embodying a potentially patentable invention before a patent application is filed, the relevant law in both the US and Europe needs to be carefully considered, taking into account the key legal differences between "selling" and "disclosing". 

In summary, please can we stop referring to G1/23 as a European on-sale bar!   

Further reading

• Confidentiality restrictions around clinical trials and prior public use (T 0670/20) (Jan 2023)
• New EBA referral: When is prior use of a product excluded from the prior art for lack of enablement? (Jun 2023)
• A more nuanced approach to the evidence standard for prior use (T 1311/21) (Jan 2025)
• Non-reproducible commercial products are prior art (G 1/23) (Jul 2025)
• [Guest Post] Has G1/23 harmoniously recast the European law of enablement? (Nov 2025)
• Non-reproducible prior art post-G 1/23: Novelty versus inventive step (T 1044/23) (Nov 2025)