IL-2023-000007 - [2025] EWHC 2863 (Ch)

Section 10(2)(b) provides, in so far as relevant that:

“(1)

…

(2)

A person infringes a registered trade mark if he uses in the course of trade a sign where because—

(a)

…

(b)

the sign is similar to the trade mark and is used in relation to goods or services identical with or similar to those for which the trade mark is registered,

there exists a likelihood of confusion on the part of the public, which includes the likelihood of association with the trade mark.”

In order to establish infringement under s 10(2), a claimant must satisfy six requirements: (i) there must be use of a sign by a third party within the relevant territory; (ii) the use must be in the course of trade; (iii) it must be without the consent of the proprietor; (iv) it must be of a sign which is at least similar to the trade mark; (v) it must be in relation to goods or services which are at least similar to those for which the trade mark is registered; and (vi) it must give rise to a likelihood of confusion: Muzmatch per Arnold LJ at [26].

Conditions (i)-(iii) are dealt with above.

Stability has conceded similarity for the purposes of (v) on Getty Images’ ‘best case’ goods and services – i.e. those expressly relied upon in closing submissions. Stability also does not contest that Getty Images can overcome the threshold requirement of similarity between Mark and Sign (condition (iv)). Accordingly an assessment of the degree of similarity becomes relevant to the question of whether “there exists a likelihood of confusion on the part of the public” (Iconix at [32]).

Given my findings above, Stability’s concession of similarity applies to both the iStock watermark* images and to the Getty Images watermark* on the First Japanese Temple Garden Image. The watermark* on the Obama Image is not representative and so I address it no further, other than to say that if I am wrong about that, then it is plainly highly similar.

I find that, even if I am wrong as to the iStock Signs on the Spaceships Image and the Dreaming Image being identical, they are nevertheless highly similar to the ISTOCK Marks. Although I have found that the Sign on the First Japanese Temple Garden Image is not identical, I find for the purposes of section 10(2) TMA that it is highly similar to the Getty Images Marks. That seems to me to be the impression that would be created on the mind of the average consumer (always bearing in mind that the average consumer rarely has the chance to make direct comparisons between mark and sign). I did not understand Stability to make any alternative submission.

The Trade Marks Registry has adopted a standard summary of the principles established by these authorities for use in the registration context, which has been approved by the Court of Appeal on a number of occasions. The most recent version of this summary was given by Arnold LJ in Muzmatch at [27], and approved by the Supreme Court in Iconix at [38]):

“(a)

the likelihood of confusion must be appreciated globally, taking account of all relevant factors;

(b)

the matter must be judged through the eyes of the average consumer of the goods or services in question, who is deemed to be reasonably well informed and reasonably circumspect and observant, but who rarely has the chance to make direct comparisons between marks and must instead rely upon the imperfect picture of them he has kept in his mind, and whose attention varies according to the category of goods or services in question;

(c)

the average consumer normally perceives a mark as a whole and does not proceed to analyse its various details;

(d)

the visual, aural and conceptual similarities of the marks must normally be assessed by reference to the overall impressions created by the marks bearing in mind their distinctive and dominant components, but it is only when all other components of a complex mark are negligible that it is permissible to make the comparison solely on the basis of the dominant elements;

(e)

nevertheless, the overall impression conveyed to the public by a composite trade mark may, in certain circumstances, be dominated by one or more of its components;

(f)

and beyond the usual case, where the overall impression created by a mark depends heavily on the dominant features of the mark, it is quite possible that in a particular case an element corresponding to an earlier trade mark may retain an independent distinctive role in a composite mark, without necessarily constituting a dominant element of that mark;

(g)

a lesser degree of similarity between the goods or services may be offset by a greater degree of similarity between the marks, and vice versa;

(h)

there is a greater likelihood of confusion where the earlier mark has a highly distinctive character, either per se or because of the use that has been made of it;

(i)

mere association, in the strict sense that the later mark brings the earlier mark to mind, is not sufficient;

(j)

the reputation of a mark does not give grounds for presuming a likelihood of confusion simply because of a likelihood of association in the strict sense; and

(k)

if the association between the marks creates a risk that the public might believe that the respective goods or services come from the same or economically-linked undertakings, there is a likelihood of confusion.”

As I have already alluded to, it is well established, and has recently been confirmed by the Supreme Court in Iconix, that it can be relevant to take into account the post-sale context when considering trade mark issues and that post sale confusion amounts to a likelihood of confusion. The rationale is that a mark may still function as a trade mark after sale because it operates as a badge of origin, and hence quality, after the goods are sold. It does so not primarily to the purchaser of the goods (who is likely to be aware of their origin), but to third parties who encounter the goods after sale (see Montres Breguet (CA) at [84]-[85] and Iconix at [75]).

In this case the Signs will be encountered after the user has downloaded the Model, or has used one of the other access mechanisms described in Appendix B. However, this does not preclude the scope for confusion. Those users will be encountering the Sign in a realistic and representative way when it appears on their screens. In some cases, the user may encounter the Sign shortly after he or she has used one of the access mechanisms. In other cases, he or she may encounter the Sign long after using one of the access mechanisms.

I have already addressed the question of the available evidence in this case. In the context of considering confusion, I bear in mind that the absence of evidence is hardly ever determinative and that claims under section 10(2) TMA can succeed in the absence of evidence.

In my judgment, the global assessment points to a likelihood of confusion on the part of a significant proportion of the relevant public as a result of the appearance of watermarks* (in the form of those appearing on the First Japanese Temple Garden Image, the Spaceships Image and the Dreaming Image) for the following reasons.

In light of these findings, I consider Getty Images to have made out its case of section 10(2) TMA infringement in relation to:

However, the above global assessment applies specifically to the three examples of use of the Sign which I have found to be identical or similar (the Spaceships Image, the Dreaming Image and the First Japanese Temple Garden Image – albeit that it is unclear which Model generated the Dreaming Image). For reasons I have explained, the analysis is highly fact sensitive – while Stability’s points on the garbled nature of the watermarks* and the nature of the images produced do not appear to me to weigh heavily in the balance on the global assessment of these three examples, it might be the case that they would do so in respect of different watermarks* generated on different images in response to different prompts. That this must be the case is borne out by a consideration of various of the examples in both Annex 8H and Annex 8I. I should add that I have not forgotten that Getty Images watermarks* have been produced by v1.x using the “news photo” prompt, but absent specific submissions directed at those watermarks* and the images on which they appear, I do not see that I can properly make a finding of section 10(2) infringement in relation to them, and (until after circulation of the draft judgment) I was not invited to do so. For reasons I have already explained, I do not consider it to be consistent with the overriding objective to permit Getty Images to have a second bite of the cherry once the judgment has been circulated.

Unlike the more usual case of trade mark infringement, it is impossible to conclude that for every watermark* generated by the same version of the Model, a similar analysis will apply such that infringement would follow automatically. The question of how many similar (infringing) watermarks* are likely to have been generated by users in the UK remains unknowable and Getty Images has advanced no case at this trial designed to address that question.