preloader logo

Legal recommendations on the use of AI in advertising campaigns

Use of advertising slogans "claims"

We discussed in our previous article, "Who Owns AI Creation?", that the use of these types of AI platforms is not without risks. No one doubts their potential and the wide range of opportunities, but after analyzing some of these platforms, such as ChatGPT, Tome, Midjourney, Runway, and EssayGenius, the first conclusion, common to all of them, is that the responsibility for the generated content rests exclusively with the user, that is, with the person who handles and enters the data (inputs), and in no case with the platform, regardless of this firm's opinion on these types of maximalist clauses.

In any case, what concerns and occupies us at this moment is that a marketing and advertising campaign can be affected, with serious economic and legal consequences, if a series of basic precautions are not taken into account.

Let's not forget that, if the campaign is designed by an advertising agency, the client understands that what is presented to them is in accordance with the law; otherwise, the agency will assume responsibility for it. No agency wants to have a campaign withdrawn, assume the costs, pay the corresponding compensation, and lose the client in the process.

Let's look at a series of examples using GPT chat, regardless of whether the agency or the client designs the campaign.

We ask GPT chat for something simple, without further details:

  • Usuario: “Give me a claim for an advertising campaign for a beef and cheeseburger.”. 
  • ChatGPT: "The burger that makes you melt with pleasure: delicious juicy meat with the creamiest cheese, try it now!"
  • Usuario: “shorter”. 
  • ChatGPT: "The best cheeseburger, savor the flavor!"
  • User: “Give me more claims like savor the flavor!"   
    Sure, here are some examples.:
    - "The flavor that makes you say 'mmm' with every bite".
    - "The explosion of flavor you were looking for".
    - "The authentic flavor that transports you to paradise".
    - "Unparalleled flavor that will leave you wanting more".
    - "The taste you will never forget".
  • User: “that the claim goes to a young audience”.
     Perfect! Here are some claim options for a young audience:
    - "Awaken your senses with the bold flavor of our cheeseburger."
    - "The funniest burger you've ever tried".
    - "The burger that ignites your passion for flavor".
    - "The cheeseburger that gives you the power to enjoy".
    - "Discover the vibrant flavor of our cheeseburger".

  • Now we ask you for ideas for an automotive campaign: 
  • Usuario: “Give me a claim for an advertising campaign for a vehicle for young people who feel free and travel.”,
  • ChatGPT: "Explore the world your way with our vehicle: the perfect companion for your adventures full of freedom and excitement.".
  • ChatGPT: "Freedom has no limits when you're behind the wheel of our vehicle: venture down every road and discover your wanderlust.".

We see that with zero intellectual effort we have a proposal for slogans such as: "The burger that makes you melt with pleasure", savor the flavor!, "The flavor that makes you say 'mmm' with every bite", "the authentic flavor that transports you to paradise", "unparalleled flavor that will leave you wanting more", "the flavor you will never forget", "awaken your senses", "the most fun burger", "the burger that ignites your passion for flavor", "the burger that gives you power", and in automotive, "explore the world your way with our vehicle", "the perfect companion for your adventures full of freedom and excitement", "freedom has no limits when you are behind the wheel of our vehicle", "venture down every road and discover your wanderlust".

In these results, where is the user's authorship? In the question "Give me a claim for a beef and cheeseburger advertising campaign?" Where is the creativity? The originality? The author's personality? The intellectual effort?

Recall, as we already mentioned in our cited article, that in our legal field, intellectual property is governed by Royal Legislative Decree 1/1996, which establishes "original creations" as the object of its protection. That said, not much more is known about what a creation is, and even less about what can be classified as original. The Provincial Court of Navarre attempted to clarify what could be understood by "creative" and "original" to determine whether or not a production will be protected by intellectual property. It gave two criteria for verifying originality: that it did not exist before and that it reflects the author's personality. Regarding creativity, the Court equates it with an "intellectual effort" that allows us to move from a simple act to an "artistic or intellectual creation." The jurisprudence of the Court of Justice of the European Union follows a similar path, more simply put. The CJEU dismissed the relevance of any consideration of the means used by the individual to protect a production, sufficing that it "reflects their personality and expresses their free and creative choices." The important thing is that human intervention exists, even if minimal, but this idea can be distorted in the examples we have presented. In any case, this debate matters little to the marketing and advertising departments, who are the ones designing the campaigns. What they want, and quite rightly, is to use generative AI for their campaigns with the lowest risk. That is why the trend toward using generative AI platforms like ChatGPT will be widespread.

We asked the latter:

User: Can you give the same result to multiple users?

ChatGPT: “As a language model, my answer to a specific question may be the same for multiple users, as long as the question is the same and the input is the same. However, answers may vary slightly due to the inherent randomness of natural language generation in models like mine. Additionally, my answers may depend on the context and prior information I have about a specific user, such as their language or interests.”.

The system recognizes that results may coincide, although we believe the likelihood of matches between users will decrease with greater intellectual intervention by the user in fine-tuning the results. In any case, the risk of a mass-market platform exists. Matches can occur. Therefore, the AI ​​platform Midjourney, which creates new images with a brief description, monetizes the risk. If you have a "Pro" account and subscribe to the "Stealth" option, Midjourney will commit to self-limiting the distribution of the result to avoid matches with third parties. This is a hint of how this business model may evolve.
Below, we will outline a series of basic legal tips that we should keep in mind when designing advertising campaigns with generative AI.

Tips: 

  1. Industrial Property. Once we have the list of selected slogans, we must conduct a search for identical or similar names or trademarks. We may find that the name we want to use is already registered, and remember that it doesn't have to be exactly the same to infringe a third party's industrial property rights. This is a task that legal departments typically perform when developing campaigns. It's important to keep in mind that an error in this regard will force you to withdraw the campaign and assume any applicable damages. Ideally, and we always repeat this, to be 100% covered, you should request registration of the claim you're interested in and not use it until you have received approval from the Registry Office. However, in practice, this seems illusory due to the speed and timeliness of advertising campaigns. We recommend at least checking that there is nothing that could cause a conflict at the time of launching the campaign.
  2. Industrial Property. When we ask the AI ​​platform for results, be careful not to make references to third-party brands, such as, "Give me a claim for an advertising campaign like [renowned trademark]," because it can give us results similar to those already registered trademarks of that renowned trademark. This must be taken into account.
  3. Intellectual property. We will address this issue in future articles because it is a highly controversial topic. The important thing when designing advertising campaigns with generative AI is that our results do not coincide with those of a third party, so it would be wise to somehow document the creation process of the result on the AI ​​platform. We would at least try to prove that there was no plagiarism if the two campaigns coincide in time.
  4. Personal data protection. Pay special attention to the personal data provided to the AI ​​platform. Remember that many of these platforms are in the sights of European data protection authorities because their headquarters are in the US and there is no information about the processing of this personal data. GDPR fines can be significant. Personal databases should never be entered under any circumstances.
  5. Confidential information. Keep in mind that the information we provide to the AI ​​platform to fine-tune the algorithm may be classified as confidential. For example, a client requests a campaign proposal from an agency. Both sign a non-disclosure agreement (NDA) where the client transmits sensitive information to the agency. Let's suppose the agency uses an AI platform and uploads the information submitted by the client to it. This would be a breach of confidentiality, given that this third party (the AI ​​platform) does not owe the agency a duty of confidentiality, especially considering that this information will remain in the system to continue improving the algorithm.
  6. Image rights. If we use results that include the name or obvious allusion to a public figure, we will likely be infringing their image rights.
  7. Review the terms and conditions of the AI ​​platform, as they will gradually change. Where rights previously belonged to the user, an updated terms and conditions may reveal that they can be shared when there is a commercial purpose.

Generative AI is undoubtedly an unpredictable tool, although we sense its great potential. It's like when, in the late 1990s, the great potential of the internet was sensed, but no one realized what would come next with social media. In the short term, there will be many legal conflicts due to the clash of legislation that doesn't fit and contradicts a new AI-based mentality. In any case, our obligation is to advise our clients so they don't violate current regulations, or at least be aware of them.

Riestra Abogados. 2023

Programa Kit Digital, iniciativa del Gobierno de España. Cofinanciado por los Fondos Next Generation EU del Mecanismo de Recuperación y Resiliencia. Kit Digital

Cookie settings

We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you.