Can you protect Artificial Intelligence inventions at the European Patent Office?

In recent years there has been a resurgence of interest in machine learning and so-called “artificial intelligence” systems. Much of this resurgence is based on advances in so-called “deep learning”, neural networks with multiple layers of connections. For example, convolutional neural networks now provide state-of-the-art performance in many image recognition tasks and recurrent neural networks have been used to increase the accuracy of many commercial machine translation systems. Machine learning may be considered a subdiscipline of “artificial intelligence” that deals with algorithms that are trained to perform tasks such as classification based on collections of data. This recent resurgence has meant that more companies wish to protect innovations in this field. This quickly brings them into the realm of computer-implemented inventions, and the nuances of protection at the European Patent Office.

A881388C-9C65-470D-AED6-9C584435DA4A
Obligatory “Terminator” Patent Attorney Stock Image

Computer-Implemented Inventions

“Computer-implemented invention” is the European Patent Office term for a software invention. Claims that specify machine learning and artificial intelligence systems are almost certainly to be considered “computer-implemented inventions”. The innovation in such systems occurs in the design of the algorithms and/or software architectures. Claims for new hardware to implement machine learning and artificial intelligence systems, such as new graphical processing unit configurations, would not be classed as computer-implemented inventions and would be considered in the same manner as conventional computer devices.


What Do We Have To Go On?

As key advances in the field have only been seen since 2010, there are few Board of Appeal cases that explicitly consider these inventions. It is likely we will see many Board of Appeal decisions in this field, but it is unlikely these will filter through the system much before 2020. However, applications in the field are being filed and examined. The following review is based on knowledge of these applications, evaluated in the context of existing Board of Appeal cases.


Prior Art

A first issue regarding machine learning and artificial intelligence systems is that many of the underlying techniques are public knowledge, given the rapid turn-over of publications and repositories of electronic pre-prints such as arXiv. Hence, many applicants may face novelty and inventive step objections if the invention involves the application of known techniques to new domains or problems. For patent attorneys who are drafting new applications, it is recommended to perform a pre-filing search of such publication sources and ensure that the inventors provide a full appraisal of what is public knowledge.


Domain of Invention

A second issue is the domain of the invention. This may be seen as the context of the invention as presented in the claims and patent description.

Inventions that apply machine learning approaches to fields in engineering are generally considered more positively by the European Patent Office. These fields will typically either operate on low-level data that represents physical properties or have some form of actuation or change in the physical world. For example, the following domains are less likely to have features excluded from an inventive step evaluation for being “non-technical”: navigating a robot within a three-dimensional space; dynamic adaptive change of a Field Programmable Gate Array; audio signal analysis in speech processing; and controlling a power supply to a data centre.

On the other hand, inventions that apply machine learning approaches within a business or “enterprise” domain are likely to be analysed more closely. These inventions have a greater chance of claim features being excluded for being “non-technical”. These domains typically have an aim of increasing profit. The more this aim is explicit in the patent application, the more likely a “non-technical” objection will be raised. For example, the following inventions are more likely to have features excluded from an inventive step evaluation for being “non-technical”: intelligent organisation of playlists in a music streaming service; adaptive electronic trading of securities; automated provision of electronic information in a company hierarchy; and automated negotiation of online advertising auctions.


Exclusions from Patentability

A third issue that arises is that individual features of the claims fall within the exclusions of Article 52(2) EPC.  In the field of machine learning and artificial intelligence systems, there is an increased risk of claim features being considered to fall into one of the following categories: mathematical methods; schemes, rules and methods for performing mental acts or doing business; and presentations of information. These will briefly be considered in turn below.

Mathematical Methods

The field of machine learning is closely linked to the field of statistics. Indeed many machine learning algorithms are an application of statistical methods. Academic researchers in the field are trained to describe their contributions mathematically, and this is required for publication in an academic journal. However, the practice of the European Patent Office, as directed by the Boards of Appeal, typically regards statistical methods as mathematical methods. In their pure, unapplied form they are considered “non-technical”.

Schemes, Rules and Methods for Performing Mental Acts

A claim feature is likely to be considered part of schemes, rules and methods for performing mental acts when the scope of the feature is too broad or abstract. For example, if a claimed method step also covers a human being performing the step manually, it is likely that the scope is too broad.

Schemes, Rules and Methods for Doing Business

Claim features are likely to be considered schemes, rules and methods for doing business when the information processing relates to a business aim or goal. This is especially the case where the information processing is dependent on the content of the data being processed, and that content does not relate to a low-level recording or capture of a physical phenomenon.

For example, processing of a digital sound recording to clean the recording of noise would be considered “technical”; processing row entries in a database of information technology assets to remove duplicates for licensing purposes would likely be considered “non-technical”.

Presentation of Information

Objections that features relate to the presentation of information may occur when the innovation relates to user experience (UX) or user interface (UI) features.

For example, a machine learning algorithm that adaptively arranges icons on a smartphone according to use may receive objections on the grounds that features relate to mathematical methods (the algorithm) and presentation of information (the arrangement of icons on the graphical user interface). As per Guideline G-II, 3.7.1, grant is unlikely if information is simply displayed to a user and any improvement occurs in the mind of the user. However, it is possible to argue for a technical effect if the output provides information on an internal state of operation of a device (at the operating system level or below, e.g. battery level, processing unit utility etc.) or if the output improves a sequence of interactions with a user (e.g. provides a new way of operating a device). Again, a technical problem needs to be demonstrated and the machine learning algorithm needs to be a tool to solve this problem.


Subfields of ML and AI

In certain subfields of machine learning and artificial intelligence, there is a tendency for Boards of Appeal and Examining Divisions to consider inventions more or less “technical”. This is often for a combination of factors, including field of operation of appellants, the history of research and traditional applications, and the background and public policy preferences of staff of the European Patent Office.

For example, machine learning and artificial intelligence systems in the field of image, video and audio processing are more likely to be found to have “technical” features that can contribute to an inventive step under Article 56 EPC. A convolutional neural network architecture applied to image processing is more likely to be considered a “technical” contribution that the same architecture applied to text processing. Similarly, it may be argued that machine learning and artificial intelligence systems in the field of medicine and biochemistry have “technical” characteristics, e.g. if they operate on data originating from mass spectrometry or medical imaging.

However, advances in search, classification and natural language processing are more likely to be found to have “non-technical” features that cannot contribute to an inventive step under Article 56 EPC. These areas of machine learning and artificial intelligence systems are often felt to be “technical” by the engineers and developers building such systems. However, it is a nuance of European case law that these areas are often deemed to have claim features that fall into an excluded “business”, “mathematical” or “administrative” category.

A recent example may be found in case T 1358/09. The claim in this case comprised “text documents, which are digitally represented in a computer, by a vector of n dimensions, said n dimensions forming a vector space, whereas the value of each dimension of said vector corresponds to the frequency of occurrence of a certain term in the document”. The Board agreed with the appellant that the steps in the claim were different to those applied by a human being performing classification. However, the Board concluded that the algorithm underlying the method the claim did not “go beyond a particular mathematical formulation of the task of classifying documents”. They were of the opinion that the skilled person would have been given the (“non-technical”) text classification algorithm and simply be tasked with implementing it on a computer.


What Should We Not Do?

Managers and executives of commercial enterprises are often habituated into selling innovations to a non-technical audience. This means that invention disclosures often describe the invention at an abstract “marketing” level. When an invention is described in a patent application at this level, inventive step objections are likely.

The fact that mathematical formulae may comprise excluded “non-technical” features is difficult for inventors and practitioners to grasp. Often equations at an academic-publication level are included in patent specifications in an attempt to add technical character. This often backfires. While such equations may be deemed “technical” according to a standard definition of the term, they are often not deemed “technical” according to the definition applied by European case law.

In general, objections are more likely in this area when the scope of the claim is broad and attempts to cover applications of a particular algorithm in all industries. Applicants should be advised that trying to cover everything will likely lead to refusal.


What Should We Do?

Chances of grant may be increased by ensuring an examiner or Board of Appeal member can clearly see the practical application of the algorithm to a specific field or low-level technical area.

Patent attorneys drafting patent applications for machine learning and artificial intelligence systems should carefully consider the framing and description of the invention in the patent specification. In-depth discussions with the engineers and developers that are implementing the systems often enable innovations to be described more precisely. Given this precision, innovations may be framed as a “technical” or engineering innovation, i.e. a technical solution to a technical problem. This increases the chance of a positive opinion from the European Patent Office.

Often features of an invention will have both a business advantage and a “technical” advantage. For example, a machine learning system that learns how to dynamically route data over a network may help an online merchant more successfully route traffic to their website; however, this improved method may involve manipulation of data packets within a router that also improves network security. A patent specification describing the latter advantage will have a greater chance of grant than the former, regardless of the actual provenance of the invention. A practitioner may work with an inventor to ensure that initial business advantages are distilled to their proximate “technical” advantages and effects. For cases where data does not relate to a low-level recording or capture of a physical phenomenon, it is recommended to ensure that any described technical effect applies regardless of the content of the data.

When considering exclusion for “mental acts”, a risk of a “non-technical” objection may be reduced by ensuring that your method steps exclude a manual implementation. Note that this exclusion does not necessarily prevent other objections being raised (see T 1358/09 above).

When drafting patent applications,  it is also important to describe the implementation of any mathematical method. In this manner, pseudo-code is often more useful than equations. It is also important to clearly define how attributes of the physical world are represented within the computer. Good questions to ask include: “What data structures and function routines are used to implement the elements of any equation?”, “How is data initially recorded, e.g. are documents a scanned image such as a bitmap or a markup file using a Unicode encoding?”,  “What programming languages and libraries are being used?”, or “What application programming interfaces are important?”.

Practitioners do need to be concerned with including overly limiting definitions within the claims; however, a positive opinion is more likely when specific implementation examples are described in the patent specification, followed by possible generalisations, than when specific implementation examples are omitted and the description only presents a generalised form of the invention along with more detailed mathematical equations.

To be successful in search, classification and natural language processing,  one approach is to determine whether features relating to a non-obvious technical implementation may be claimed. This approach often goes hand in hand with a knowledge of academic publications in the field. While such publications may disclose a version of an algorithm being used, they often gloss over the technical implementation (unless the underlying source code is released on GitHub). For example, is there any feature of the data, ignoring its content, which makes implementation of a given equation problematic? If inventors have managed to reduce the dimensionality of a neural network using clever string pre-processing or quantisation then there may be an argument that the resultant solution is implementable on mobile and embedded devices. Reducing a size of a model from 3 GB to 300 KB by intelligent selection of pipeline stages may enable you to argue for a technical effect.


Do Not Believe The Hype?

Despite the hype, machine learning and artificial intelligence systems are just another form of software solution. As such, all the general guidance and case law on computer-implemented inventions continues to apply. A benefit of the longer timescales of patent prosecution is that you ride out the waves of Gartner’s hype cycle. In fact, I still sometimes prosecute cases from the end of the dotcom boom…

Computer-Related Board of Appeal Decisions from 2014 Case Law Summary

The European Patent Office has now published a case law summary for 2014 as Official Journal supplementary publication 4/2015. Sections I-C-4.1, I-C-5.1, II-A-1, and II-E-1.4 discuss case law that relates to computer-related inventions. The relevant passages are extracted and commented on below.

Skilled Person

T 407/11 held that the relevant skilled person in the context of providing computer-system users with operating assistance via a user interface (e.g. error messages or warnings) was an expert in software ergonomics concerned with the userfriendliness of human-machine interfaces rather than an expert in software programming or in computer technology in the strict sense.

The objective problem to be solved by that skilled person was to prevent a situation whereby the user’s action caused an electronic data-processing system to execute a called function differently from intended (or even to fail to execute it at all).

In the board’s view, however, the technical effect claimed in the application (simpler operation of an object-oriented user interface, facilitating initial use and subsequent familiarisation, especially for beginners or upgrading users, and so making the resulting method easier and more intuitive to learn) could not be considered a directly derivable consequence of the distinguishing features, because attributes such as “simpler operation” or “easy and intuitive familiarisation” were generally subjective, i.e. depended on the user’s individual preferences and experience or intellectual capabilities, while the classification of users as “beginners”, “upgraders” or “advanced” was generally based on a variety of criteria which were not clearly defined.

This suggests that providing objective definitions of technical effects (e.g. millisecond time savings) may help to support an inventive step under European practice. It also further indicates a need to avoid reference to “user”-based advantages.

Applications of Algorithms

In T 2035/11 the application mainly related to navigation systems that could be tailored to a user’s particular wishes. The focus of the application was on the route-planning functionality of a navigation system.

The board held that the subject-matter of claim 1 lacked an inventive step within the meaning of Art. 52(1) and 56 EPC. It noted that mathematical algorithms may contribute to the technical character of an invention only in so far as they serve a technical purpose (see e.g. decision T 1784/06). The purpose of the algorithm was the mere display of an optimal path to the user for cognitive processing. The user could act on the information, but did not need to.

As stated in decision T 1670/07, a technical effect may arise from either the provision of data about a technical process, regardless of the presence of the user or its subsequent use, or from the provision of data (including data that on its own is excluded, e.g. produced by means of an algorithm) that is applied directly in a technical process.

In the case at issue, the data was produced by means of an algorithm and was not applied directly in a technical process, so neither possibility applied.

Warning on Generalisation

In T 2231/09 the patent in suit concerned a method of representing and analysing images. Claim 1 of the main request set out that “… at least one said descriptor element is derived using only a subset of pixels in said image.”

The board considered the expression “subset of pixels” to be problematic under Art. 84 EPC 1973 and stressed that, while a certain degree of generalisation may be permitted, features as claimed should make it possible to clearly identify features of embodiments that are covered by the terms of a claim. Moreover, the generalised subject-matter as claimed should make it possible to understand the technical problem to be solved.

When amending claim 1, the applicant had put forward a new interpretation according to which a “region” could mean the whole image, and a “subset” could correspond to all pixels of the region. The board considered this interpretation to be inconsistent with essential parts of the described embodiments, according to which a subset corresponded to only some of the pixels of a region. The subject-matter of claim 1 was thus unclear when interpreted in the light of the description.

The board also stated that the requirements of clarity and support by the description in Art. 84 EPC 1973 were designed to reflect the principle that the terms of a claim should be commensurate with the invention’s technical contribution to the art. Taking into account the description, the board regarded the division of the image into regions and subsets as essential for achieving the technical effect underlying the invention. Therefore, the subject-matter of claim 1 was not supported by the description. The board concluded that claim 1 did not comply with Art. 84 EPC 1973.

This indicates the need, when drafting claims for computer-related inventions, to provide clear and unambiguous definitions of terms used within the claims. This is especially important when features of the claims relate to abstract entities, e.g. data within a data processing system.

Added Subject Matter and Features without Technical Contribution

In T 1779/09 the board considered that the appellant had found itself exactly in the situation envisaged in decision G 1/93 (OJ 1994, 451). As emphasised in Headnote II of G 1/93, “a feature which has not been disclosed in the application as filed but which has been added to the application during examination and which, without providing a technical contribution to the subject-matter of the claimed invention, merely limits the protection conferred by the patent as granted by excluding protection for part of the subject-matter of the claimed invention as covered by the application as filed, is not to be considered as subject-matter which extends beyond the content of the application as filed within the meaning of Art. 123(2) EPC.” These principles were confirmed in G 2/98 (OJ 2001, 413) and G 2/10 (OJ 2012, 376). The board in the case at issue concluded that a limiting feature which generally would not be allowable under Art. 123(2) EPC could, under certain conditions, nevertheless be maintained in the claim of an opposed patent in the particular situation addressed in decision G 1/93. It then complied with Art. 123(2) EPC by way of a legal fiction. In the case at issue, the term “only” was introduced during the examination proceedings and successfully objected to under Art. 100(c) EPC in proceedings before the opposition division by the former respondent. Since the board considered the term to be truly limiting, its deletion would extend the protection conferred and thereby infringe Art. 123(3) EPC. However, the board held that the exclusive limitation did not influence the solution of the technical problem as understood from the application as originally filed, and hence provided no technical contribution to the claimed invention (see also decision T 384/91, Headnote II). It merely excluded protection of part of the invention described in the application, thus not giving any unwarranted advantage to the applicant. Claim 1 of the appellant’s sole request was therefore deemed to comply with Art. 123(2) EPC.

Case Law Review – T 2296/10

Case:

T 2296/10

Claimed Subject Matter:

OFDM-transmitting apparatus and method, and OFDM-receiving apparatus and method.

Comments:

This case provides an interesting consideration of technical prejudices.

The Appellant attempted to argue that technical prejudices of the company behind the closest prior art document (the BBC) meant that the claimed solution was not obvious.

The Board disagreed with the Appellant’s arguments stating (see 2.1.10):

In this regard, the board notes that, in general, cost considerations and technological preferences of a particular company (like the BBC in this case) cannot impose technical prejudices or uncertainties upon a technically skilled person such that he would be deterred from envisaging a technically sound and feasible solution for that reason alone. Otherwise, when analysing and interpreting the actual teaching of the closest prior art determined for the purpose of assessing inventive step according to the well-established “problem-solution approach”, internal experiences, beliefs, and preferences concerning technologies to be applied by the company from which that closest prior art originates would generally have to be taken into account. This would in turn mean that – to answer the question whether the skilled person starting out from the closest prior art would in fact arrive at the claimed solution – additional internal background information on the respective closest prior art (e.g. derived from witness statements from some employees as provided by the present appellant) would be necessary. In other words, the extent to which the notional skilled person in fact applies his skills in providing a solution to an objective technical problem would be unduly bound by such internal information (e.g. the expected infrastructure costs incurred by the BBC when proposing a change in its applied technology) at the filing date of an application instead of finding an appropriate solution to the objective problem posed. That would, however, definitely be incompatible with the problem-solution approach as generally applied.

Rather, the person skilled in a technical field (i.e. mobile communication networks in the present case) would try to seek a technical solution to an objective technical problem (i.e. choosing a certain pilot symbol pattern) under certain constraints (i.e. digital video distribution), starting with the closest prior art (i.e. document D1). However, based on the reasoning set out in points 2.1.12.1.1 to 2.1.92.1.9 , the board considers that said skilled person would arrive at the solution (i.e. using a DVB-T-based pilot symbol pattern) according to feature A) of claim 1 without exercising inventive skills.

[With thanks to Jake Loftus for help finding and reviewing these cases.]

Case Law Review – T 1602/09

Case:

T 1602/09

Claimed Subject Matter:

A  computer system for managing relationships between brokers and traders in a trading network.

Comments:

The Appellant claimed that the invention solved a technical problem which arose in such systems, namely that a trader could not simply and easily control, i.e. prevent or permit, a computer terminal operated by a broker to send trading commands on behalf of the trader from the computer terminal via the network to the trading system.

The Appeal was dismissed by the Board.

The Board found that the problem formulation only made sense in the context of operations by a (non-technical) “active trader”, these being traders who wish to be able to supervise trade orders given to brokers. It was found that since the trading was computer-based the active trader would need to have access to the broker’s trading system. The skilled person was certainly aware that this made a log-on necessary. Equivalently, if the active trader for some reason was not logged on, the broker should not be allowed to trade. Ideally, the check should be automatic. These straight-forward considerations were deemed to lead directly to the subject-matter of claim 1.

[With thanks to Jake Loftus for help finding and reviewing these cases.]

Case Law Review – T 1192/10

Case:

T 1192/10

Claimed Subject Matter:

User interface with gesture recognition.

The subject-matter of claim 1 differed from the disclosure in the closest prior art document in that according to claim 1:

a) warning messages are output to the user instead of having only internal messages between different components,

b) a range of sampling periods, i.e. a period when the button is activated, is supervised,

c) a range of poses indicating ranges of pitch and roll angles of the input device with respect to the bottom plane on which the input device is positioned is supervised and

d) temporary button release when the button is deactivated and re-activated within a predetermined time causes another warning to the user.

Comments:

Feature (a) of outputting warning messages to a user instead of internal messages between different components was deemed to be technical, solving the technical problem of how to provide feedback to the user about internal states or identified conditions of the device, but notoriously known.

However, features (b) of supervising a range of sampling periods and (d) providing temporary de-activation of a button were deemed to be technical features that provided an inventive step over the cited art. This set aside an earlier decision of the examining division.

The examining division had argued that although the actual implementation of the specific validation criteria involved a technically skilled person, the definition of the expected operation (e.g. what motions and durations of the input are expected) was rather business-based, according to the intended purpose of the device and to design choices. This argument was not accepted by the Board who concluded that, whatever the reason for the definition of a gesture might be, the underlying ranges, rolls and angles are of a technical nature.

Case Law Review – T 0913/10

Case:

T 0913/10

Claimed Subject Matter:

Encrypting data within a database system based on a defined system role (security administrator, database administrator, and user administrator).

Comments:

The case provides an interesting discussion about the term “role”, e.g. as used in a computing sense e.g sysadmin, user, database admin.

The Board concluded that “that nothing in the claims or in the description can dispel the reasonable possibility that the definition of tasks to be distributed over three administrators is merely an organisational and hence non-technical issue, not­withstanding that it relates to a technical entity such as a database system”. The claims were then found to lack an inventive step based on this assumption.

[With thanks to Jake Loftus for help finding and reviewing these cases.]

Case Law Review – T 1929/09

Case:

T 1929/09

Claimed Subject Matter:

Server-side web summary generation and presentation.

Claim 1 specified: an Internet Portal, comprising an Internet-connected server and a portal software executing on the server, including a summary software agent, wherein the Portal maintains a list of Internet destinations specific for a subscriber, and the summary software agent accesses the Internet destinations, retrieves information according to pre-programmed criteria, and summarizes the retrieved information for delivery to the subscriber.

Comments:

The underlying idea was to facilitate the life of a user who normally had to provide some personal information for accessing certain web pages. This was deemed at first instance to be a non-technical problem.

The Board agreed with the Appellant that improving the allocation of resources on the Internet and reducing connection time between servers or between a server and a workstation are, in principle, technical problems. However, the Board held that an information service tailored to the needs of a particular user and limited to user-defined Internet destinations is essentially a business scheme. Thus, the mere idea of providing such a service cannot be regarded as a contribution to the solution of a technical problem.

Claim 1 was broad and lacked technical detail on a specific implementation; as such any technical features therein were deemed to be obviously required. The background of the invention was cited to support this point.

[With thanks to Jake Loftus for help finding and reviewing these cases.]

Software & Patenting: IP Outside Your Comfort Zone

A presentation given as a CIPA Webinar on 25 February 2014.

Provides an introduction to software as it relates to patenting and an overview of current practice in UK and Europe. Details of relevant legislation and case law are provided, together with some tips for drafting.

Provided according to the terms set out here: http://www.eip.com/legal.php – i.e. does not constitute legal advice and should be taken as guidance.

Case Law Review – T 0218/11

Case:

T 0218/11

Claimed Subject Matter:

A self-service checkout which solved the problem of the self-service checkout being overly sensitive (or conversely not sensitive enough) to people making mistakes (or conversely, trying to “cheat” the checkout).

Comments:

The Appellant argued that keeping track of customers and tolerating different numbers of errors when using the checkout was itself technical, and that this would form part of the problem for the technically skilled person to solve.

However, the Board concluded that judging whether a customer is trust-worthy and treating them according to that judgement was a non-technical matter. Hence, an underlying idea of recording a level of trust forms part of a requirements specification that is given to the skilled person; the technically skilled person is faced with the task of modifying the self-service checkout terminals so as to keep track of how trusted different customers are, and so as to interrupt transactions earlier for those customers who are less trusted, and later for those that are more trusted.

The features of the claimed solution was thus deemed to either be found in the prior art, be non-technical and thus not contribute to an inventive step, or be technically obvious given the defined technical problem.

[With thanks to Jake Loftus for help finding and reviewing these cases.]

Case Law Review – T 2216/09

Case:

T 2216/09

Claimed Subject Matter:

A a system that enabled subscribers of a wireless telecom operator to execute financial transactions with a mobile phone.

Comments:

The system was deemed to mainly relate to an excluded business scheme.

All steps of the underlying business scheme were deemed part of the information provided to the technician in charge of the technical implementation and did not as such contribute to inventive step.

The Appellant argued that the specific transaction platform and client software did not exist in a conventional wireless phone system, and so were out of reach of the normal activity of the person skilled in the art of telephone networks. However, the Board concluded that the person skilled in the art would be able to implement the new system, given the specifications of the underlying business scheme. For example, any extension of the type of financial transactions which can be performed with the account (receive monetary deposits, debit and credit operations) was deemed to be dictated by the underlying business scheme.

[With thanks to Jake Loftus for help finding and reviewing these cases.]