Interim guidance from the UK IPO on computer-implemented inventions

The Supreme Court has overturned years of practice in the UK in examining computer-implemented inventions. 

While the Emotional Perception case was about the patentability of an artificial neural network, the Justices took the opportunity to reject the Aerotel test laid down in 2006 by the Court of Appeal and used by the UK Intellectual Property Office ever since. The UK IPO has issued unofficial interim guidance to assist patent attorneys and applicants during this period of upheaval.

This note should be read alongside our overview of the UK Supreme Court’s landmark judgment, “Landmark UK Supreme Court judgment reshapes patent law for AI and software inventions.”

Why has the Supreme Court made this decision? #

Patenting computer-implemented inventions #

Patent offices around the world have long wrestled with the issues of patenting computer software. Patent law in most countries specifies subject matter that is excluded from patentability, such as software. However, laws generally do not specify a clear boundary between what is excluded subject matter and what is an invention, and it is for the courts to determine a suitable test. In the UK, the recent rise in the number of AI-related applications has pushed the Aerotel test to its limit.

The Aerotel test #

For the last twenty years, the UK Intellectual Property Office has applied the Aerotel test to determine whether a claimed invention is directed to patentable subject matter. Crucially, this test was carried out before considering the novelty and inventiveness of the claims, and indeed, the IPO would refuse to search an application that failed it. However, the test required the step of identifying the contribution made by the invention, which necessitates an understanding of the prior art. This paradox made the test difficult to apply consistently.

This difficulty was highlighted by the Supreme Court, which noted that while the IPO, the High Court and the Court of Appeal all considered the Emotional Perception claims using the Aerotel steps, they reached different conclusions.

In the years since the Aerotel decision, the practice of the European Patent Office has gone in a different direction. In the Comvik method set out by the Enlarged Board of Appeal in case G1/19, a claim overcomes the patentability hurdle if it includes any hardware at all. Then, only the technical features of the claim, i.e., those that are not excluded from patentability, are taken into consideration for inventive step.

This test works on the principle that while many patentable inventions are implemented by computers, it is not inventive to program a computer to simply carry out a task that is already done by hand, or to perform a non-technical method, such as an unapplied mathematical method, a business method, or the creation of an aesthetic design. Rather than refuse such applications for containing excluded subject matter, the EPO refuses them for being non-inventive.

The Supreme Court unanimously decided that UK practice needs to follow a similar “any hardware” approach, as anything else is inconsistent with the European Patent Convention on which UK patent law is based.

A new UK test #

While it might seem like an answer would be to import the whole of the EPO’s Comvik test, this uses the “problem-solution” approach that is not found in UK practice. It is therefore likely that the “any hardware” approach will be combined with the UK’s Pozzolli test for inventive step. This is a four-step process:

1. Identify the background knowledge of a “person skilled in the art”;
2. Identify the inventive concept of the claim;
3. Identify the differences between the prior art and the inventive concept; and
4. Determine whether those differences constitute steps that would have been obvious to the skilled person.

If an approach similar to Comvik were applied here, then step 2 or 3 (or both) could be carried out based only on technical features of the claim. However, the Pozzolli test suffers from a similar malady as Aerotel, namely, identifying the inventive concept before considering the prior art, and therefore, it is not yet clear what path the IPO will take. 

Interim guidance #

It will take time for the IPO to determine the new test, to train examiners, and to update the Manual of Patent Practice. In the meantime, interim unofficial guidance has been issued via the Chartered Institute of Patent Attorneys. All of the below applies only to applications that have been considered under Aerotel.

Examination reports #

It is still necessary to respond to Examination Reports produced under the Aerotel test. The IPO has requested that attorneys do not request fresh examination reports but respond using the approach of “firstly identifying an invention (applying the ‘any hardware’ criterion) and secondly identifying claim features and their interactions that contribute to the technical character of the claimed invention”. This is not only a stopgap measure but will provide the IPO with useful feedback on how attorneys are interpreting the judgment.

Hearings #

All hearings for applications rejected under the Aerotel test are cancelled and will be rescheduled after an opportunity for further written communications.

Search reports #

While the IPO will now search all applications that pass the low “any hardware” hurdle, existing applications where a search was refused will not be automatically included, and a request will need to be filed in such cases. The IPO asks that this be done only for applications within the twelve-month convention period, where it will make the most difference, but it is an option for all applications. 

Extensions of time  #

Due to the ongoing uncertainty, the IPO indicates that requests for extensions of time to respond to examination reports are likely to be granted, and likewise extensions to the compliance period (although the fee will not be waived).

Missed deadlines #

In cases where applicants did not request examination by the deadline due to a refusal to search, there is a two-month extension available, but the IPO indicates that further discretionary extensions will not be automatically agreed.

For lapsed applications, a change of practice at the IPO is not an accepted reason for restoration. However, it might be possible to show “a continued underlying intention to proceed” if a decision was made not to respond to an unfavourable examination report.

Other considerations #

Unduly limited claims #

On some applications, limitations may have been introduced to overcome the Aerotel test that are now not necessary. Applicants and attorneys may wish to review all pending cases to identify whether an alternative route might be taken under a different test.

For applications that are close to grant, rather than unpicking the existing prosecution, a divisional application could be filed.  

Lapsed applications #

Under IPO practice, an application remains pending until the compliance period expires even if a deadline for responding has been missed, and therefore the records of applicants and attorney records may contain a number of applications marked as lapsed that are actually pending. If a response was not filed because of difficulties overcoming an objection based on the Aerotel test, it may be worth revisiting the application as the IPO is likely to allow an extension of time, even years later.

What this means in practice #

This shift represents a significant change in how computer-implemented inventions are assessed in the UK, particularly for AI and software-related technologies. Applicants should consider reviewing pending and lapsed applications in light of the new approach, as opportunities may now exist that were previously unavailable under the Aerotel framework.

As the UK IPO refines its position, a proactive and flexible prosecution strategy will be key to navigating this evolving landscape.