When Does Importation of Refurbished and Refilled Printer Ink Cartridges Infringe a Patent?

If you have purchased a consumer-grade printer at any time during the past decade or so, then you will be well aware that the manufacturers of these products have mostly adopted the so-called ‘razor-and-blades’ business model.  According to this model, an initial item – such as an inkjet printer – is sold at a low price (possibly below cost) in order to increase subsequent sales of consumable supplies – i.e. ink cartridges.  Unsurprisingly, an entire industry has evolved to supply cheaper ‘generic’ versions of these consumables, either by manufacturing copies or, more commonly, by refurbishing and refilling used cartridges.  Printer manufacturers, in their turn, have developed a range of strategies in an effort to thwart generic competition and maintain an effective monopoly on the sale of consumables.

For example, some of the measures employed by original equipment manufacturers (OEMs) are physical, such as forming uniquely-patterned structures (‘interface patterns’) on the exterior of ink cartridges so that they can only be installed into the specific printer models for which they are intended.  Other strategies involve more sophisticated technological measures, such as configuring memory chips installed in the ink cartridges – which perform a legitimate function in monitoring ink usage and level to prevent damage to print heads that results from operating the printer with an empty cartridge – to frustrate efforts to reset the memory contents.  If the memory chips cannot be reset or reproduced by a generic supplier, then even a refilled cartridge will fail to function with a printer that interrogates the memory and finds an indication that it is empty.

Intellectual property rights also play a role in the ongoing struggle between OEMs and generic suppliers.  In some jurisdictions (though not, it must be said, Australia) registered design rights can be used to protect the physical appearance of cartridges – including ostensibly functional interface patterns – and thus prevent the manufacture and sale of copies.  Functional improvements in cartridge technology can also be protected by patents, again enabling the sale of copies to be restrained through infringement actions.

However, refurbishing and refilling of cartridges represents a grey area for IP laws.  The default legal position is that once a product (which the common law would class as a ‘chattel’) has been sold, the purchaser becomes the owner, and is free to deal with it as he or she wishes, including reselling, donating, or disposing of it, whereby any subsequent owner may enjoy the same rights of possession.  This raises the question of whether there can be infringement of an IP right, such as a patent, when instead of making and selling a copycat product, a competitor legally acquires spent ink cartridges and restores them to (roughly) their original condition for resale?  The answer – as I am sure any lawyer will tell you – is ‘it depends’.

A decision issued by Justice Burley in the Federal Court of Australia addresses a particularly interesting variation on this theme.  Seiko Epson Corporation v Calidad Pty Ltd [2017] FCA 1403 concerns the collection, refurbishment, and refilling of ink cartridges for Epson-branded printers outside Australian jurisdiction, and subsequent importation and sale of the refilled cartridges by Calidad.  While the case covers a range of issues, including trade mark infringement, trade practices and contract law, the most substantial aspect of the decision – and the one of greatest interest to this blog, of course – is whether or not importing and selling the refurbished cartridges constitutes infringement of Seiko’s Australian patents covering certain features of its cartridges.

The short answer, according to Burley J, is ‘it depends’!  In particular, it depends upon exactly what is done in the course of refurbishment, and whether this results in the product being ‘materially altered’, insofar as it is ‘an embodiment of the invention as claimed.’  If you want to know what that means, and how his Honour arrived at this unique conclusion, read on.

A Purchaser’s Rights in a Patented Product

Under earlier Australian (and UK) legislation, a patent gave the patentee an exclusive right to ‘make, use, exercise and vend the invention’.  The current Australian Patents Act 1990 grants a more comprehensive right to ‘exploit’ the invention which, in the case of a product, includes to ‘make, hire, sell or otherwise dispose of the product, offer to make, sell, hire or otherwise dispose of it, use or import it, or keep it for the purpose of doing any of those things’.  Clearly, this encompasses the old rights, while making clear that such other acts as importing and warehousing products embodying a patented invention also fall within the patentee’s exclusive rights.

Notably, these rights are in conflict with the common law notion of ownership of a chattel – absent some additional, unstated, limitation, the exclusive right to ‘use’ a product embodying a patented invention means that upon sale or resale of a product the patentee will have continuing control of the use.  Two competing legal theories have arisen to address this apparent problem:

1. the doctrine of ‘exhaustion of rights’, which posits that upon a first authorised sale of a product by or with the approval of a patentee, all patent rights in the subject matter of the invention are exhausted; and
2. the notion that there exists an implied licence to use a product in any manner in which the purchaser desires upon purchase, unless the patentee imposes conditions at the time of sale.

The exhaustion of rights approach has been adopted in the United States, as was recently confirmed by the Supreme Court in Impression Products, Inc. v Lexmark International, Inc. 137 S. Ct. 1523.

The position in Australia is, arguably, less clear.  The most influential authority remains a 1911 decision of the Privy Council, in National Phonograph Co of Australia Ltd v Menck [1911] UKPCHCA 1; (1911) 12 CLR 15, which was decided under the Patents Act 1903, and which – as Burley J most eloquently explains at [82]-[92] of Seiko Epson – essentially adopts the implied licence approach.  More recent authority on the matter is limited to obiter dicta in a copyright case (Interstate Parcel Express Co Pty Ltd v Time-Life International (Nederlands) BV [1977] HCA 52), some UK decisions (Roussel-Uclaf v Hockley International Ltd [1996] RPC 441 and HTC Corporation v Nokia Corporation [2013] EWHC 3247), and a 2016 decision of a single judge of the Federal Court of Australia including the caveat that ‘[n]otwithstanding the changes in the patent and copyright statutory regimes, I propose to proceed upon the basis that National Phonograph continues to be the law as it applies to standard patents and innovation patents, and that the decision in Interstate still applies to copyright material’ (Austshade Pty Ltd v Boss Shade Pty Ltd [2016] FCA 287, at [82]).

Justice Burley’s reasoning in Seiko Epson proceeds on the basis – as in Austshade – that National  Phonograph continues to apply, and that the rights of subsequent owners of products embodying patented inventions are protected by an implied licence.

Scope of the Implied Licence

National Phonograph and its progeny also establish that a patentee may impose conditions on the initial sale of products embodying a patented invention, thereby limiting the scope of the implied licence.  However, such restrictions do not ‘travel with the goods’ – any subsequent owner receives the full benefit of the implied licence unless that owner has actual knowledge of the restrictive conditions at the time of acquiring the goods.  Obligations to ensure that knowledge of the restrictions is ‘passed on’ can, of course, be imposed by contract (which is not too far removed from what the US Supreme Court found to be the case in the context of exhaustion of rights, in Lexmark). 

Patentees wishing to impose downstream restrictions on the exploitation (including use, export/import, and resale) would be well-advised to read the summary of propositions set out by Burley J at paragraphs [110]-[117] of Seiko Epson.  Fortunately, we do not need to delve into the potential complexities of all this here, because the court found that, in any event, neither the company Ninestar (which performed the actual refurbishment and refilling of cartridges), nor Calidad (which imported the resulting cartridges for sale in Australia) acquired the products subject to any restrictions imposed by Seiko (at [127]-[143]).

Importantly, his Honour found that ‘there is no difficulty implying the licence where the patented goods are first sold outside of the patent area (that is, broadly, outside Australia)’ (at [117]).  Thus, in the absence of any restrictions, it is irrelevant that the cartridges may not originally have been sold into the Australian market, or that Seiko may or may not hold equivalent patent rights in the country of original sale, or in any subsequent jurisdiction through which the products may have passed – the implied licence nonetheless subsists.

Rights to Refurbish or Refill?

The ‘traditional’ position regarding the repair of patented products is legally simple: an owner has a right to maintain and repair a product that they own in order to prolong its life.  However, assessing this in a particular case may be factually tricky – at some point a repair or refurbishment may become so extensive that the owner is effectively making a replacement product, which is, of course, an exclusive right of the patentee not encompassed by the implied licence. 

Whereas in the UK ‘repair’ was once considered to be covered in this way by the implied licence, the position in recent UK authorities is that the right to ‘repair’ is not one of the exclusive rights granted to a patentee, and therefore no licence is required (Seiko Epson at [146]-[151]).  The only question, then, is whether what is done amounts to more than repair, and encroaches upon the patentee’s exclusive right to ‘make’ the invention.

Either way, whether an owner’s actions amount to a permissible ‘repair’, or an unauthorised ‘making’, is a matter of degree.

Can Modification ‘Extinguish’ the Implied Licence?

The difficulty for Seiko Epson, however, is that regardless of whether refurbishing and refilling printer cartridges amounts to a ‘repair’ or ‘making’ of the patented products, the acts were not performed in Australia, and therefore could not amount to an infringement of the Australian patents in any event.  Rather, the already refurbished products are imported and sold in Australia.  They embody the patented inventions, and therefore infringe the patents unless their importation and sale is covered by an implied licence.

The question is therefore not whether the refurbishment is an infringement of the patents, but rather whether modifications performed on the cartridges in the course of refurbishment and refilling extinguish the implied licence that undoubtedly attached to the original sales.  As stated by Burley J, at [163]:

The proper question is accordingly not based on a repair or making analysis, but rather whether the implied licence can be said to apply in circumstances where one owner of the product, Ninestar, has modified it and then passed title in it to [Calidad]. The relevant question then becomes; at what point has an owner of the embodiment made changes to it such that the implied licence arising from a sale sub modo is extinguished? No authority appears to have directly addressed this topic.

‘Material Alteration’ – A New Concept in Patent Law?

The ‘threshold question’, according to Burley J, is ‘whether or not the modified product is materially the same embodiment of the invention as claimed as the product that the patentee sold without restriction.’  More particularly, since the scope of the patentee’s rights is defined by the patent claims, this notion of modification must be tied to the invention as claimed.  That is, ‘[t]he question is not whether or not the product was altered or repaired, but whether the product, insofar as it is an embodiment of the invention as claimed, was materially altered, such that the implied licence can no longer sensibly be said to apply’ (Seiko Epson at [164], italics in original).

In simple terms, a modification that is unrelated to the patent rights such as, for example, painting the product a different colour (assuming colour forms no aspect of what is claimed), would not amount to a material alteration, and would not extinguish the implied licence.  However, removing and/or replacing a component of the product that is actually an element of the invention as claimed may be a material alteration that would extinguish the implied licence.

As Burley J has explained it (at [178]):

This process of reasoning leads to the conclusion that to assess whether or not the implied licence continues after modifications are made one must consider the significance of modification work done on a product and how the modifications in question relate to the features of the patented product that are defined by the claim. Where that work done, or alteration made, does not concern a claimed feature, then the work is irrelevant to the analysis. That is because the patentee’s rights to limit the use of the patented product arise because the product represents an embodiment of the claimed features.

In that same paragraph, his Honour went on to summarise a three-stage factual inquiry for making the assessment (because there is, of course, nothing better in the law than a three-step test):

1. determine the scope of the invention as claimed;
2. assess the manner in which the patentee’s product is an embodiment of the invention as claimed; and
3. determine the extent to which the modifications made affect the patented product insofar as it represents an embodiment of the claims.

The Claimed Invention

By agreement between the parties, the decision focuses on a single claim, namely claim 1 of Australian patent no. 2009233643.

The patent relates to printer cartridges that contain a memory chip which operates from a low voltage (typically 3.3V) and a sensor device that requires a higher voltage (typically in excess of 30V).  Contacts on a printed circuit board integrated with the cartridge provide for the application of low- and high-voltage power supplies from the printer, as well as for communications between the printer and the memory chip.  The object of the invention is to reduce the risk of damage to the cartridge and/or printer as a result of short-circuiting between low- and high-voltage terminals.

The claimed solution comprises 11 integers, which are set out in full at paragraph [200] of the decision.  It is sufficient for present purposes to reproduce a shorter summary from paragraph [201]:

[T]he features of the product claimed, broadly, are that it is an ink cartridge that can be attached to a printer by inserting it, with both the ink cartridge and printer having a plurality of terminals. The ink cartridge must have a memory driven by a memory driving voltage (integer [2]), an electronic device driven by a higher voltage (integer [3]), terminals electrically connected to the memory and terminals electrically connected to the electronic device (integer [4]), and contact portions for the terminals which are arranged in a particular way (integers [5]-[11]). (Bolding in original.)

‘Material Alteration’ of Printer Cartridges

A number of different refurbished and refilled products were said by Seiko to have infringed the patent, distinguished on the one hand by the particular models of printer with which they were intended to operate, and on the other hand – and more relevantly to the issues in the case – by the particular steps carried out in order to refurbish and refill the cartridges.  There were, in fact, a number of distinct processes of reverse-engineering, reprogramming, and physical modification of cartridges, combined in various ways for different products, however the features that were ultimately found to be relevant can be reduced to:

1. products having memory chips that were reset (by re-writing their contents using specialised equipment) without physical replacement;
2. products having memory chips that were replaced, by removing the entire printed circuit boards (including the chips and terminals/contacts) from the cartridges, removing the chips and replacing them with compatible devices, before refitting the modified circuit boards to the cartridges (noting that the circuit boards would not generally end up back on the same cartridge from which they were removed – see paragraph [271] of the decisions); and
3. products including interface patterns to restrict compatibility to particular models of printer, where the modifications include physical removal of the interface pattern.

For products having memory chips that were electronically reset without replacement, or other physical modification of the cartridges, the court found that there was no ‘material alteration’.  In particular, the information content of the chip, which is changed by this modification, forms no part of what is claimed as the invention, even though the chip itself is recited in the claim (see, e.g., paragraph [243]).  Thus, resetting the content of the chips does not result in the implied licence being extinguished, and importing and selling this category of refurbished and refilled cartridges does not infringe Seiko’s patents.

For the other categories of products, however, the court found that the modifications did involve a ‘material alteration’ such that the implied licence was extinguished.  Replacing the memory chip – which is an explicit integer in the patent claim – was a material change to the embodiment of the invention as originally sold by Seiko (see, e.g., paragraph [276]).

With regard to the interface pattern, the claim specifically recites ‘a printing material container adapted to be attached to a printing apparatus by being inserted in an insertion direction’.  The court found the interface pattern to be ‘part of the mechanism by which Seiko’s embodiment achieved that function’ and thus that, while possibly borderline, removing the pattern fell on the wrong side of the line such that this modification would extinguish the implied licence (at [281]).

Overall, therefore, the outcome was mixed for the parties on both sides.  While a number of categories of refurbished and refilled products, sold prior to April 2016, were found to lack the benefit of the implied licence, and therefore to infringe the Seiko patents, other categories of products sold since that date were found to be non-infringing by virtue of the continuing existence of the implied licence.

Conclusion – Innovative Reasoning, Ripe for Appeal?

On a personal note, I have to say that I love this judgment.  It is a joy to read, and provides an elegant exposition of the history of the law in relation to the implied licence attached to patented products, and the right to repair.

I also find Justice Burley’s reasoning (mostly) compelling.  But there is no question that it involves development of the law in order to address a novel situation that has not previously been considered in any relevant judicial authority.  There are a number of unsettled and/or innovative elements to this reasoning, as summarised in the bullet points below, and a failure of any one of these would be likely to alter the outcome.

1. The decision depends firstly on it being the law in Australia that the rights of a subsequent owner of a patented product to deal with it as they see fit are based upon an implied licence, rather than the alternative proposition of an exhaustion of the patentee’s rights upon first sale.  There remains a question as to whether statutory changes since 1911 – and particularly enactment of the Patents Act 1990, with its broad concept of the right to ‘exploit’ a patented invention – may have altered the position.
2. Secondly, the court has developed what appears to be a new concept in patent law, namely the notion that the implied licence (assuming it exists) can be extinguished by a ‘material alteration’ to the product.  To be clear, this is something that is to be distinguished from the established ’repair or making’ analysis (Seiko Epson at [163]).
3. Thirdly, assessment of whether or not a patented product has been ‘materially altered’ involves an inquiry into the significance of the modification work done on the product and how the modifications in question relate to the features that are defined by the patent claims.  Modifications that do not concern a claimed feature are irrelevant to the analysis.  However, modifications that operate directly on claimed features, such as removing and/or replacing a component of the product that is actually an element of the invention as defined in a claim, may result in ‘material alterations’.  This is something different from the usual ‘whole of claim’ infringement analysis since, by definition, the product both before and after modification falls within the claim.

In view of this, and the fact that the parties on both sides have some reason to be dissatisfied with the outcome, it seems to me that there are good prospects that the decision will be appealed.  And I think that would be a good thing.  There is an opportunity here for a Full Bench of the Federal Court (and possibly, ultimately, the High Court) to settle whether ‘implied licence’ or ‘exhaustion of rights’ is the correct approach in Australia, and to confirm or correct the reasoning developed by Burley J in this judgment.  Personally, I would like to see it upheld but, more importantly, it would be beneficial to have clarity on the law either way in relation to refurbishment and reselling of patented products.

Tags: Australia, Exhaustion, Implied licence, Patent law