AI in web design contracts

The use of artificial intelligence (AI) in web design and programming opens up many new possibilities. As a web designer or programmer, you can use AI to implement customer orders more efficiently and innovatively. But what do you need to consider contractually when creating AI-generated content or designs for customers? Here are the most important legal aspects for your web design contracts.

The use of AI tools such as ChatGPT, DALL-E or Midjourney is tempting, as they can generate impressive texts, images and designs in seconds. However, this is a legal gray area. This is because there are still no specific laws that comprehensively regulate the use of AI. Nevertheless, web designers and programmers must comply with applicable law, particularly with regard to copyright, rights of use, liability and data protection.

With advanced AI systems such as Sonnet from Anthropic’s Claude.ai, the possibility of web designers and programmers no longer having to write any code themselves is even within reach. Sonnet can create websites independently based on descriptions in natural language, including HTML, CSS and JavaScript. This opens up completely new perspectives for the industry, but also raises questions about the future of the profession. One thing is clear: AI will fundamentally change the way we develop websites.

Copyright and rights of use

A key issue is the question of copyright for AI-generated content. According to the current legal situation in Germany, the copyright generally lies with the developer or user of the AI, not with the AI itself. According to copyright law, only personal intellectual creations of a person can be protected by copyright. Simply operating an AI is usually not enough. Only if the human gives the AI such specific instructions that the design of the work is already largely determined could copyright protection arise.

As a web designer or programmer, you should always ask yourself the still partially unresolved legal questions of whether and when you can transfer rights to AI-generated content to customers at all, which you may not even hold yourself through the use of AI. This has a significant impact on the drafting of contracts with customers.

Contracts should therefore regulate very carefully which rights of use are granted to the customer for the designs and content created using AI. The scope and duration of the rights of use must be precisely defined. Even if there is no copyright on the AI results themselves, so-called ancillary copyrights may arise, e.g. for the creator of a database.

In addition, when using AI outputs, it must always be checked whether the rights of third parties are infringed, for example if the AI was trained with copyrighted material. This is a gray area with legal risks.

In order to create the greatest possible legal certainty, it is advisable to transfer all conceivable types of use of the AI-generated content to the client as comprehensively as possible in contracts. These include, in particular, the right of reproduction, distribution, exhibition, presentation, performance and demonstration, the right of broadcasting, the right of making available to the public and the right of adaptation and transformation. The granting of rights of use should be unlimited in terms of territory, time and content.

At the same time, however, service providers should also make it clear in their contracts that they are not liable for any legal infringements caused by the AI. After all, you have no influence on the training data and how the AI works. This risk should be distributed fairly between the contracting parties.

Of course, this only applies if web designers or programmers are transparent about the fact that they use AI to create orders. If, on the other hand, you conceal the use of AI from the customer, you are not only operating in a legal gray area, but also risk considerable liability risks. In this case, you cannot claim that you cannot fully understand and control how the AI works.

Instead, you give the customer the impression that this is a purely human service for which you are fully responsible. Any errors or infringements by the AI are then the sole responsibility of the service provider. In addition, most customers are likely to see it as deception if they receive AI-generated content without their knowledge. This can lead to a loss of trust and reputational damage.

From a legal perspective, it is therefore strongly recommended to disclose and contractually regulate the use of AI when creating orders. This is the only way to distribute risks and responsibilities fairly and avoid unpleasant surprises. Transparency creates trust – concealment, on the other hand, harbors considerable legal and business risks.

Contractual arrangements for the commercial use of AI outputs are therefore a balancing act with many uncertainties. A careful risk assessment is necessary in each individual case. Until there is a clear regulation by the legislator, we are operating in a legal gray area. Web designers and programmers should be aware of this and draft contracts accordingly. If in doubt, it is advisable to consult a lawyer specializing in IT law.

Liability and warranty

Despite the use of AI, you as a web designer or programmer remain responsible to the customer for the results. Do not contractually exclude your liability for errors or infringements by the AI. Check AI-generated content carefully before handing it over to the customer. Also make it clear that you are liable for the technical and graphical integration of the content, but not for content-related statements that the AI makes independently.

After all, as the contractor, you are the one who selects, configures and uses the AI systems. You must ensure that the tools used are suitable and sufficiently trained for the respective application. Even if the AI generates content independently, you are responsible for ensuring that this output complies with quality and legal requirements. To minimize liability risks, you should know the limits and risks of the AI systems used and communicate them transparently to the customer. Make explicit reference in your general terms and conditions and project contracts to the fact that you use AI tools and what consequences this may have. Also clearly define your area of responsibility and separate yourself from areas that are outside your control. It is also advisable to actively involve the customer in the AI-supported creation process. Ask for interim results, obtain regular feedback and adjust the AI parameters accordingly.

The closer you work with the customer, the lower the risk that you will end up with an unsatisfactory result for which you will have to take responsibility.

Despite all precautionary measures, liability for AI-generated content can never be completely ruled out. The technology is still too immature and unpredictable for that. As a web designer or programmer, you should be aware of this and always plan for a risk buffer. Also check whether your business or property liability insurance covers damage caused by the use of AI. This means that you are at least financially protected in the event of an emergency.

Data protection and confidentiality

If you use the customer’s personal data to train your AI systems, you need a legal basis for this. Obtain the customer’s consent and provide transparent information about the type, scope and purpose of data processing. You should also contractually ensure that the customer only provides you with content and data for which they have the necessary rights. Always treat all customer information confidentially.

The customer’s consent must be given voluntarily, for the specific case, in an informed and unambiguous manner. It should take the form of a declaration or other unambiguous confirmatory act. Silence, boxes that have already been ticked or inactivity on the part of the data subject therefore do not constitute consent. The consent must explicitly refer to the processing of personal data for AI training purposes. A blanket declaration of consent for all processing purposes is ineffective.

Inform the customer clearly and comprehensibly about the planned data processing. This includes information on the controller, purposes of processing, categories of personal data, recipients or categories of recipients, duration of storage or criteria for determining the duration, information on the rights of data subjects and the existence of a right of appeal to a supervisory authority. You can include this information in the privacy policy on your website, for example.

Make sure that the customer only provides you with data for which he has the necessary rights of use. This is particularly relevant if the customer in turn passes on personal data of third parties, e.g. its own customers or employees, to you. In this case, the contract should contain a guarantee from the customer that they have all the necessary rights and consents for the transfer of the data to you.

Always treat the data received from the customer as strictly confidential and protect it from unauthorized access using suitable technical and organizational measures. Create a processing directory and carry out a data protection impact assessment if necessary. Involve the company data protection officer in the planning of AI projects at an early stage.

Please also note the rights of the data subjects whose data you process. This includes in particular the right to information, rectification, erasure, restriction of processing, data portability and objection. Set up processes to fulfill these rights within the legal deadlines.

If you act as a processor within the meaning of Art. 28 GDPR, you will conclude a data processing agreement with the customer. In particular, the object and duration of the processing, the type and purpose of the processing, the type of personal data and the categories of data subjects are defined therein. The obligations and rights of the controller are also regulated.

Careful contractual regulations and transparent communication with the customer can minimize many data protection risks when using AI. Nevertheless, it remains a challenge to fully transfer the requirements of the GDPR to self-learning systems. A precise examination and consideration is required in each individual case. In case of doubt, it is advisable to consult the competent data protection supervisory authority.

Conclusion: Drafting contracts in the age of AI requires IT nerds with legal expertise

AI is revolutionizing the work of web designers and programmers. However, the new possibilities also come with legal pitfalls. To avoid this, contracts are needed that are tailored to the special challenges of AI. This is the only way to minimize liability risks, secure copyrights and rights of use, guarantee data protection and maintain confidentiality.

But who can draw up such contracts? Experts are needed who understand both the technical and legal aspects of AI down to the last detail. IT nerds with legal expertise, so to speak. Because only those who know how AI systems work and what data they process can make the right contractual arrangements.

Lawyers who can program themselves and are familiar with the latest technologies have a clear advantage here. They can bridge the gap between technology and law and develop customized solutions. Because they understand the language of developers as well as the language of lawyers.

As a lawyer with years of experience in the IT sector and a passion for technology, I bring precisely this combination to the table. I know the challenges facing web designers and programmers in the age of AI and can support them in drafting legally compliant and future-proof contracts.

Because one thing is clear: AI will fundamentally change the way we develop websites and software. If you want to be successful here, you need to be up-to-date not only technically, but also legally. Continuous training at the interface of AI and law is therefore essential.

With a strong partner at their side who is equally proficient in the language of technology and law, web designers and programmers can make the most of the opportunities offered by AI – and keep the risks under control. So nothing stands in the way of a successful future in the age of AI.

Author: Marian Härtel

Marian Härtel ist Rechtsanwalt und Fachanwalt für IT-Recht mit einer über 25-jährigen Erfahrung als Unternehmer und Berater in den Bereichen Games, E-Sport, Blockchain, SaaS und Künstliche Intelligenz. Seine Beratungsschwerpunkte umfassen neben dem IT-Recht insbesondere das Urheberrecht, Medienrecht sowie Wettbewerbsrecht. Er betreut schwerpunktmäßig Start-ups, Agenturen und Influencer, die er in strategischen Fragen, komplexen Vertragsangelegenheiten sowie bei Investitionsprojekten begleitet. Dabei zeichnet sich seine Beratung durch einen interdisziplinären Ansatz aus, der juristische Expertise und langjährige unternehmerische Erfahrung miteinander verbindet. Ziel seiner Tätigkeit ist stets, Mandanten praxisorientierte Lösungen anzubieten und rechtlich fundierte Unterstützung bei der Umsetzung innovativer Geschäftsmodelle zu gewährleisten.