HT-2021-000363 - [2025] EWHC 532 (TCC)

Mr Anzani - 2018

Mark Anzani is Special Projects Executive at IBM Corp. His responsibilities include monitoring potential competitors or potential partners to IBM’s mainframe technology. On his own description, his mode of operation is a bit “cloak and dagger”. He freely admitted that typically he does not write a great number of things down and purposefully does not keep too many records of some of his actions; others within the IBM Group are informed only on a “need to know basis”.

Mr Anzani’s evidence was that he first heard about LzLabs in about February 2016, when an industry contact informed him that LzLabs was expected to announce the SDM product at the forthcoming CeBIT 2016 computer conference.

At the CeBIT conference on 14 March 2016 LzLabs publicly launched the SDM. Mr Anzani stated in cross-examination that he became very interested and immediately carried out initial investigations into LzLabs:

“Q. And almost immediately you carried out some investigations into LzLabs yourself, didn't you?

A. That, yes, is when I really took interest in -- in LzLabs.

Q. And you became aware, extremely early on, that the ultimate owner of LzLabs was Mr John Jay Moores?

A. Yes.

Q. And you knew, because of your involvement in the litigation involving Neon, who he was?

A. Yes, that is correct.

Q. You also carried out investigation of who else was involved in relation to the matters and you discovered the existence of Texas Wormhole?

A. Yes, I know -- I know, during the enquiries, Texas Wormhole's name came up. I don't recall exactly when in -- in the process.

Q. Very early on. Very early on.

A. I -- as I recall, it was -- it was relatively early, yes. Exactly when, I don't know, but it was early, yeah, I'll give you that.

Q. And you also came to understand that there had been certain individuals who had transferred to Zurich in relation to the LzLabs development; correct?

A. Yes, that is correct.”

From LzLabs’ publicly available marketing materials, Mr Anzani learned that LzLabs claimed that the SDM would be able to run IBM mainframe applications in their object-level form without changes or compromise to performance. On about 25 April 2016, Mr Anzani set up a formal investigation into LzLabs and the SDM through a special project named “Project Eiger”.

Mr Anzani described the project as an investigation to understand the technical capabilities of the SDM that was kept separate from any marketing strategy by IBM Corp’s subsidiaries. In cross-examination he accepted that there were enquiries from clients of IBM, IBM France, IBM Germany and IBM Switzerland, to which he responded. This included, on occasions, discouraging those clients from moving their applications to LzLabs. When it was put to him that from summer of 2016 he decided upon a strategy which he maintained until 2019, namely, that he would try and put customers off, but would not institute any legal proceedings, Mr Anzani stated:

“ No, I would say that that's not correct, or at least I have to be more clear and put some context around it. Before instituting any legal proceedings, you need to have facts and technical information about the solution and sufficient depth of understanding to know that, you know, if legal proceedings are going to be decided to be taken by -- by IBM, that you have a basis upon which to do it. So there was the period of time, several years, where very little information is available and we continued to work to try to reveal and understand what we could. In parallel with all of that work, I was working to do my best to ensure that customers had a set of questions that I felt were relevant, with the idea hopefully that they would continue to stay with IBM and not choose to go with the SDM.”

In September 2016, a two-day conference was held in Toronto for the engineers to discuss theories as to how the SDM might work. Mr Anzani stated that during this period, technical reports were produced within Project Eiger but they were not shared with others in the IBM Group.

Mr Anzani’s evidence is that he did not know about Winsopia in 2017. He said that he suspected that LzLabs must have access to a mainframe but he had no information as to if, or how, they had such access. He became aware of a link between LzLabs and Winsopia in July 2018, when IBM Corp’s legal team were contacted by Robert Soprano, a former employee of LzLabs and former employee of IBM Corp, who told them that he thought that LzLabs was stealing from IBM and identified the involvement of a subsidiary company called Winsopia. This information was passed on to Mr Anzani.

On 16 July 2018, Mr Anzani requested Marcy Nechemias, an IBM z data analyst, to carry out a search for Winsopia in the IBM Corp global database. Although an earlier search for LzLabs in October 2017 had produced no hits, the 2018 search by Ms Nechemias indicated that LzLabs was listed as Winsopia’s parent company in the database.

Subsequently, Mr Anzani arranged a call with Mr Soprano in September 2018, during which Mr Soprano disclosed that LzLabs was working in conjunction with Winsopia to reverse engineer IBM’s code in order to steal IBM’s intellectual property.

Thus, by July 2018, there was evidence that amounted to a trigger, putting Mr Anzani on a course of enquiry which led to the discovery of the breaches.

Mr Anzani’s evidence is that throughout his initial discussions with Mr Soprano, he did not discuss or disclose the link between Winsopia and LzLabs to any employee at IBM. To the best of his recollection, the first time he discussed Winsopia with IBM was in late 2020 at the time of the Winsopia audit requests.

In cross-examination, he stated that it was not until 2020 that the Project Eiger team gained an understanding of how the SDM worked, including details of the data structures that are set up as a part of the compilation process, which are not publicly available but would be necessary for a link-edited application in its binary form to operate on a different technical platform. It was this insight that led the team to believe that this must have been achieved through use of IBM’s intellectual property.

The defendants’ position is that the knowledge of Mr Anzani (and, if necessary, members of Project Eiger, including Mr Mitchell) should be attributed to IBM, applying the general principles of agency as set out in Meridian Global Funds Management Asia Ltd v Securities Commission (above) per Lord Hoffmann at pp.506-507:

“The company's primary rules of attribution will generally be found in its constitution, typically the articles of association … There are also primary rules of attribution which are not expressly stated in the articles but implied by company law …

…The company … builds upon the primary rules of attribution by using general rules of attribution which are equally available to natural persons, namely, the principles of agency. It will appoint servants and agents whose acts, by a combination of the general principles of agency and the company's primary rules of attribution, count as the acts of the company. And having done so, it will also make itself subject to the general rules by which liability for the acts of others can be attributed to natural persons, such as estoppel or ostensible authority in contract and vicarious liability in tort.”

In Bilta v Nazir (above) Lord Mance stated at [41]:

“As Lord Hoffmann made clear in Meridian Global, the key to any question of attribution is ultimately always to be found in considerations of context and purpose. The question is: whose act or knowledge or state of mind is for the purpose of the relevant rule to count as the act, knowledge or state of mind of the company?”

This issue was considered, albeit in a different factual context, by the Privy Council in Julien v Evolving Tecknologies and Enterprise Development Co Ltd [2018] UKPC 2. In that case, the issue was whether knowledge of a sole shareholder should be attributed to the company for the purpose of ascertaining the date of knowledge under the relevant limitation legislation, in circumstances where claims were made against former directors for negligence that was said to have been deliberately concealed. Having accepted the well-established principle of company law and rule of attribution, namely that the unanimous decision of all the shareholders in a solvent company about anything which the company under its memorandum of association has power to do should be the decision of the company, Lord Briggs considered the difficulties inherent in extending that principle, so as to attribute to the company a sole shareholder’s knowledge of, or the ability to discover, a breach of duty owed to the company by its directors:

“[52] Section 14 of the Limitation Act is concerned not with the knowledge of claimants at a particular moment in time when taking some positive step, but rather with their knowledge, or their means of discovering the relevant facts, exercising due diligence, within some period after the occurrence of the breach giving rise to a cause of action.

[53] This is not a concept easily applicable to a body of shareholders, who have no reason to be unanimous about matters concerning the company otherwise than when making some relevant decision which, by virtue of their unanimity, is treated by the law as an act of the company. It is one thing to say that shareholders making a unanimous decision may have the type of knowledge, constituting an intent that the thing resolved upon should be done, sufficient to be attributable to the company as mens rea. It is quite another thing to say that shareholders, who may or may not be unanimous when asked to make a decision about the company, should have attributed to the company the knowledge of some of them, or even all of them, about the facts relevant to a wrong done to the company by its directors, or the ability, with reasonable diligence, to discover them during some period of time.

[54] Some, but not all, of these difficulties may be surmountable where the company has a sole shareholder, rather than a body of shareholders who may, or even frequently do, act unanimously. The problem that they may have different views from time to time plainly disappears, at least where the single shareholder is an individual or a corporation sole. But a number of other problems remain. First, the ordinary basis upon which the knowledge of directors or agents of a company is attributed to the company is that they owe a duty to the company to report relevant knowledge about its affairs. In sharp contrast, it is a cardinal principle of company law that shareholders do not owe such duties to their company. Shareholders are, in principle, entitled to leave their company to make its own inquiries about its affairs and, in particular, owe no duty of reasonable diligence to inform themselves about facts which might give rise to a claim by the company against wrongdoers, even against its directors.

[55] The absence of any such duty was a powerful factor leading the courts below to reject the more generally based submissions about the attribution of shareholder knowledge. Mr Knox submitted that, in the limitation context, that absence of such a duty on even a sole shareholder made no difference, because the policy behind the Limitation Act was concerned with the prosecution of stale claims by those with an interest in them, rather than by those with a duty to sue. But this submission misses the point, because it fails to respect the separate identities of the company and its sole shareholder. True it is that the company need have no more than an interest in bringing proceedings for s.14 to be applicable to it. But s.14 applies to the company as potential claimant rather than to its shareholder. The attribution question is nonetheless powerfully affected, in this context negatively, by the absence of any duty of the shareholder to report its, or his, knowledge to the company.

[56] For these and other reasons it is difficult to treat the attribution to a company of the knowledge of its own shareholders about facts relating to a claim by the company against its directors as a general rule of attribution. The general rule is that it is the knowledge of the company’s directors that is attributed to it and, in appropriate cases, the knowledge of its agents.”

IBM’s position is that it is only where a person acts on behalf of a company (whether as director, employee or some other form of agent) that their knowledge is capable of being attributed to the company. It is said that, as a matter of statutory construction, the knowledge of an agent is not attributable to the company for the purposes of section 32.

Reliance is placed on the decision in: Peco Arts Inc v Hazlitt Gallery [1983] 1 WLR 1315 at 1326 E-G, in which Webster J said that the acts or omissions of an agent of a claimant are not to be attributed to the claimant for the purpose of ascertaining any lack of reasonable diligence, having regard to the express words in section 32(1):

“References in this subsection to the defendant include references to the defendant’s agent and to any person through whom the defendant claims and his agent.”.

The issue in that case was whether the claimant, acting through its art expert agent, had used reasonable diligence in not having had a painting authenticated and therefore, the judge in that case did not consider directly the issue whether the agent’s knowledge should be attributed to the company. However, in Allison v Horner [2014] EWCA Civ 117, a broader application was expressly accepted as correct by Aikens LJ at [15]:

“… in Peco Arts Inc v Hazlitt Gallery Ltd Webster J held that the acts or omissions of an agent of the claimant were not to be attributed to the claimant for the purposes of section 32(1). Thus knowledge of the deceit alleged on the part of a claimant’s agent will be insufficient to start the limitation period running under section 32(1). Similarly, the fact that the claimant’s agent could with reasonable diligence have discovered the alleged deceit does not start the limitation period running. I would accept this construction of section 32(1) for the reasons that Webster J gives at page 202G-H of the report.”

Further, approval of this approach was given in Riyait v Dawett [2018] EWCA Civ 593 at [34]-[37] per Sir Geoffrey Vos C (where the knowledge of the claimants’ solicitors was not attributable for section 32 purposes):

“[34] … In Horner this court held that only knowledge of the defendant's agents may be attributed under section 32, so that "knowledge of the deceit alleged on the part of a claimant's agent will be insufficient to start the limitation period running under section 32(1)", and "the fact that the claimant's agent could with reasonable diligence have discovered the alleged deceit does not start the limitation period running", and "the knowledge of agents of Mr Horner concerning the fraudulent representations is not to be attributed to him".

[35] … It is clear that the qualifying words in section 32(1) apply to the whole of the subsection, so that the attribution of the knowledge of the defendant's agents applies in all parts of the subsection, just as the non-attribution of the knowledge of the claimants' agents applies in all parts of the subsection.”

The defendants submit that Riyait, Horner and Peco Arts are authority for the proposition that the knowledge of external (and arm’s length) entities/individuals such as solicitors, tax consultants and valuation houses cannot be attributable to a claimant for the purposes of section 32. They are not authority for the proposition that for the purposes of section 32 the knowledge of an individual (or indeed other group entities or group task forces) through whom the company is operating (whether as employee/director or in a role analogous to that) is not to be attributed to the company.

The defendants’ submission is correct. As the above authorities make clear, the court must ascertain the date by which IBM, and not its external agents, discovered, or could with reasonable diligence have discovered, the concealment for the purpose of section 32(1). However, IBM is a corporate entity, not a natural person. When considering the state of IBM’s knowledge, the question that arises is whether the knowledge of any employee or agent should be attributed to it under the established rules of attribution set out in Meridian and Bilta. As the authorities make clear, the answer to the question whether an individual’s knowledge should be attributed to a company depends on context and is fact sensitive.

In this case I find that Mr Anzani was not a servant or agent of IBM when carrying out his investigations as part of Project Eiger and his knowledge should not be attributed to IBM. Firstly, Mr Anzani was employed by IBM Corp and has never been an officer or employee of IBM. Secondly, Mr Anzani had no involvement in the ICA between IBM and Winsopia, prior to the audit request in December 2020. Thirdly, Mr Anzani was not appointed by IBM to investigate LzLabs and the SDM; his work for Project Eiger was undertaken in his role as employee of IBM Corp. Fourthly, Mr Anzani had no obligation to disclose his investigations or report to IBM. His evidence was clear that it was his practice to keep secrets from everyone else in the IBM Group and that individuals were given information on a “need-to-know” basis. Fifthly, Mr Anzani’s evidence, which I accept, was that he did not tell anyone at IBM of his knowledge about the link between LzLabs and Winsopia prior to 2020.

Mr Anzani admitted in cross-examination that he acted on behalf of the whole IBM Group, including IBM, when orchestrating the message to be disseminated to existing or potential customers approached by LzLabs or interested in the SDM:

“Q. You are the executive, aren't you, who is on point and has been on point since February of 2016, trying to protect IBM's interests against the competitive threat of LzLabs?

A. Yes, I agree with that.

Q. And you've done that on behalf of IBM UK and on behalf of IBM Corporation?

A.

Yeah, I have been on point to protect our interests across all of the IBM companies.”

A distinction must be drawn between the acts carried out by Mr Anzani on behalf of the wider IBM Group, to protect the Group’s market share, and knowledge acquired by Mr Anzani during his investigation into LzLabs and the SDM, through Project Eiger, as an employee of IBM Corp. That latter investigation was not carried out as employee or agent of IBM.

For the above reasons:

There is a dispute between the parties as to the date on which any damage occurred so as to give rise to a cause of action in the tort of unlawful means conspiracy. The competing dates are 2016 (the defendants) and 2021 (IBM). The evidence before the court on the date of damage is sparse, not least because issues of quantum were not included in this trial. I do not consider that it is necessary to resolve this matter because on the facts it does not affect any limitation defence or the key findings on liability. The only claim that might be affected would be IBM’s claim against Winsopia for unlawful means conspiracy, which might be caught by the contractual limitation period of two years, subject to the issue of dishonest or deliberate concealment. However, based on my Judgment dated 29 November 2023, when giving permission to IBM to amend to plead the unlawful conspiracy claim, I am satisfied that, as a result of late disclosure, IBM could not with reasonable diligence have discovered the concealment regarding the unlawful conspiracy claim prior to June 2023, when that claim was formulated.

It follows that (subject to the disputed date of damage which, on the basis of my other findings, does not affect the outcome), the claims are not contractually time-barred or statute-barred for limitation.