(No) Copyright Protection for Logos Generated by AI

On 13 February 2026, Munich Local Court handed down a much-noted judgment on the possible – in this case, denied – protection of works created using generative artificial intelligence (Case No.: 142 C 9786/25).

The subject of the ruling was the defendant’s use of the following logos generated by generative AI.

The plaintiff had generated the logos using AI, having provided it with instructions for creating the logos, some of which were detailed and some repetitive.

The plaintiff sought an injunction and removal on the basis of Section 97(1) of the German Copyright Act (UrhG). The plaintiff argued that the logos he had created by means of prompting were works protected by copyright.

The court dismissed the claim.  The logos were not copyright-protected works of applied art within the meaning of Section 2(1)(4) and (2) of the German Copyright Act (UrhG).

Concept of a work

In its decision, the court first addressed the concept of a work under Section 2(2) of the UrhG and EU law. The concept of a work is an autonomous concept under EU law and comprises two elements: it must be an original work in the sense that it constitutes an intellectual creation of its own in the mind of its author. Furthermore, this creative content must be expressed in the work.

Originality exists only if the object reflects the author’s personality by expressing their free creative decisions.

If, on the other hand, the design is determined by technical considerations, rules or other constraints that leave no room for artistic freedom, the necessary originality is lacking.

Application to AI-generated content

The court applies these requirements to the present case. Whether AI-generated products have the character of a work depends, in the court’s view (referring to the prevailing view in the literature), on the extent to which human creative influence is still exercised despite the software-controlled process.

Copyright protection is conceivable if a human intervenes in AI results – even retrospectively or successively during the prompting process – and their personality is reflected in the output. Nevertheless, human creative influence on the design of the specific work itself is required, for example through sufficiently individualised settings or through a creative selection process; the mere selection of a suggestion from among several AI results is not sufficient.

If the generation is entirely software-controlled, copyright and related rights in the AI-generated work are excluded (para. 19).

Requirements for human contribution

The decisive factor for the court in assessing the logos at issue was whether the plaintiff’s prompting expressed his creative abilities in an independent manner by making free and creative decisions and thereby also imparting his personal touch to the output.

The use of the AI model must “be closer to an aid than to an independent creative instrument”. In the court’s view, this is only the case if the creative elements incorporated into the prompting dominate the output to such an extent that the object as a whole can be regarded as the human being’s own original creation.

General, open-ended instructions in which the AI makes the creative decisions are insufficient, even if these instructions are numerous or carefully formulated.

With regard to the logos in dispute, the court stated that the claimant had described the logo featuring a laptop, with a book bearing a paragraph symbol floating in front of its screen, to the AI only by means of a brief, two-line description of a “simple but unusual” logo for a legal text website. In the case of the logo featuring an envelope in front of a building with columns, the claimant had indeed formulated a prompt of approximately 1,700 characters, which the court classified as elaborate; however, it essentially viewed this as merely a time-consuming exercise without any independent creative creation.

Similarly, in the case of the logo “two people of different skin colours shaking hands with a bell”, the court assessed the initial prompt (“Create a logo for a career & jobs notification application … use the shape of a handshake and a bell icon … adapt the handshake and bell shapes to form something unique and creative”) as not going beyond the description of a brief given to a human designer.

In the court’s view, many of the specifications were so general (“modern, minimal, distinctly original”, “clean flat design with custom geometric abstraction”) that they did not allow any conclusions to be drawn about the specific appearance of the logo. In some instances, the claimant deliberately left the selection and precise design of the elements to the AI (“waves, motion lines, rays … if you deem them a good fit”), thereby assigning the creative decision to the AI.

The court also largely classified the plaintiff’s subsequent interventions (correction of obvious AI errors, adjustment of skin tones, feminisation of the hand, etc.) as technical, craft-based activities that do not constitute an original expression of his personality.

In the court’s view, the technical activity of the AI predominates when viewed as a whole; there is a lack of sufficient human creative influence. The court thus denies that the plaintiff’s personality has been given creative expression in its own right and, consequently, that a work eligible for copyright protection exists.

Practical relevance of the judgment

The Munich Local Court sets out detailed criteria for when AI-generated content may or may not acquire the status of a work protected by copyright.

The decision makes it clear that mere ‘prompting’, even in an extensive and technically sophisticated form, is not sufficient if the actual creative decision is left to the AI. Where the human contribution is minimal or merely technical and technical in nature, there is strong evidence that the logo is not protected by copyright – nevertheless, other intellectual property rights such as trade marks and designs must be examined separately.