You are using an outdated browser and your browsing experience will not be optimal. Please update to the latest version of Microsoft Edge, Google Chrome or Mozilla Firefox. Install Microsoft Edge

November 27, 2023

Navigating Legitimate Use of Intellectual Property in Generative AI

Subscribe to Updates

The emergence of generative artificial intelligence (AI) has transformed the landscape for innovators and creators. As many legal practitioners have pointed out, it’s imperative for both developers of AI and artists using generative AI to understand the intricacies of intellectual property (IP) strategies so they can navigate this evolving terrain successfully. This article lays out some essential considerations relating to the major types of IP for both developers and creators in the realm of generative AI.

IP Strategies for Developers of Generative AI

Developers of generative AI technologies play a pivotal role in the innovation landscape. There are three overarching IP-related issues to consider: protecting their own intellectual property, mitigating the risk of violating other people’s IP rights, and IP commercialization. Key aspects of these concerns, along with suggested approaches for developers, are outlined below.

Protecting IP

  • Copyrights. One of the primary considerations for AI developers is the protection of AI-generated works, such as art and source code. The good news is that in most countries, these creations enjoy copyright protection without the need for registration. As a result, the works are automatically protected from the moment of creation. However, it’s crucial to maintain comprehensive records of your work to establish your ownership.
  • Trademarks. Trademarks are vital for AI developers looking to establish and protect their brand. Pay close attention to Nice classifications, particularly class 9 (for software), class 35 (for business management and online marketing), and class 42 (for software design and development). Registering trademarks in these classes can provide robust protection for your brand and products.
  • Patents. For truly innovative AI algorithms, techniques, or processes, consider the option of patenting. Patents offer strong protection, but they require a thorough application process and the documentation of your innovation, including evidence that the invention is novel, non-obvious, and practically applicable.
  • Trade secrets. Protect valuable assets such as training datasets and algorithms through trade secrets. Keep them confidential and restrict access to maintain their secrecy. Legal agreements, including nondisclosure agreements (NDAs), can be invaluable in safeguarding these critical components.

Mitigating the Risk of Infringing Others’ IP Rights

  • Data collection. When collecting datasets for training AI models, ensure that you use legally obtained data. Commissioned datasets and orphan works (copyright-protected works for which the author cannot be identified or found) often come with clear usage rights, making them safer choices. Exercise caution when dealing with datasets of uncertain origin to avoid potential infringement issues.
  • Data protection laws. AI training datasets may contain personal data, which can be subject to data protection laws such as the General Data Protection Regulation (GDPR) in Europe or the Personal Data Protection Act (PDPA) in Thailand. Noncompliance with applicable data protection laws such as these can harm individuals’ privacy rights and risk legal consequences.

IP Commercialization

  • Ownership issues. Ownership of IP can become complex, especially when multiple parties are involved in creating it. To avoid future disputes and legal complications, be diligent in clarifying ownership rights and potential liability issues.
  • Licensing components. Licensing IP rights is a fundamental tool for exploiting them commercially. The license should also define the scope of ownership over the work, including elements created independently by the author and those developed through the use of generative AI. When commercializing your AI-generated products, identify the components that you own or have created. Distinguishing between your own IP and components that incorporate open-source or others’ licensed IP can make it easier to craft effective licensing agreements while protecting your interests.

IP Strategies for Creators and Artists

Creators and artists using generative also face a range of IP challenges. While some of these relate to the broader topics identified above, the specific concerns and recommended approaches differ. For instance, creators and artists have their own unique concerns when it comes to IP commercialization and the risk of infringing IP rights. The issue of IP ownership, particularly in relation to derivative works and jointly created inventions using generative AI, also comes to the fore.

IP Commercialization and Licensing

  • Licensing and selling. Creators and artists using generative AI have the opportunity to license or sell their works if the end-user license agreement permits. This can provide a pathway for monetization and sharing your creative endeavors with a broader audience.
  • Ownership of IP rights: Determining ownership rights is a necessary prerequisite when considering the commercialization of IP created through generative AI and prompts. Factors such as the “sweat of the brow doctrine,” which considers the level of user contribution to the generated work, can influence ownership. Carefully examine the terms of use and licensing agreements for any AI tools and platforms used in creating a work. If explicit ownership terms are absent, the local laws in each jurisdiction may determine ownership.
  • Commercial arrangements. The type of service model, such as licensing or subscription, can also impact copyright ownership rights. Whether it’s a free or paid subscription, be aware of their implications on your ownership when using a third party’s AI software to create your works, as the ownership rights may be influenced by the chosen licensing or subscription models.

Ownership

  • Derivative works. When creating derivative works using generative AI, it may be necessary to obtain authorization from the original sources, especially if the original works are copyright protected.
  • Jointly created inventions. In situations involving jointly created inventions or when using generative AI as a collaborative tool, clearly defining ownership and responsibilities among collaborators (e.g., via agreement) can prevent disputes and ensure that all contributions are acknowledged and protected.

Risk of Infringing Others’ IP Rights

  • Utilizing trusted generative AI. To mitigate the risk of infringing others’ IP rights or breaking other laws (such as the GDPR), always use generative AI with legally sourced materials, including datasets and pretrained models.
  • Exercise caution with fair-use doctrine. Fair use requirements and exceptions can vary significantly across jurisdictions. What qualifies as fair use in one place may not be in another, so it’s necessary to understand the local legal context and be cautious about relying on fair use as a defense.

Conclusion

Generative AI presents both exciting opportunities and complex challenges for innovators, developers, creators, and artists. Legal complexities—particularly those related to IP—are prominent among the implications of this technology, so effective IP strategies are essential for safeguarding creations, avoiding legal pitfalls, and capitalizing on the benefits of innovation. Careful and well-thought-out legal IP strategies will help creators, innovators, and their organizations make meaningful progress in this dynamic landscape.

Related Professionals

Industries

Technology

Practices

Intellectual Property

Locations

RELATED INSIGHTS​ 

August 25, 2026
Vietnam has enacted a new decree establishing administrative penalties for violations in the fields of cybersecurity and personal data protection. Decree No. 330/2026/NĐ-CP (Decree 330), issued and effective from August 19, 2026, provides a detailed sanctions framework for noncompliance with the Law on Personal Data Protection (including its implementing regulations under Decree 356/2025/ND-CP) and the Law on Cybersecurity, together with their guiding decrees. The issuance of Decree 330 signals that the practical grace period previously perceived by many businesses may be drawing to a close, with active regulatory enforcement in these areas expected to commence in earnest. Scope and Key Provisions Decree 330 has extraterritorial effect and applies to both onshore and offshore companies. For offshore companies, it applies to those that (1) provide telecommunications, internet, online-content, information-technology, cybersecurity, or cross-border services and (2) are involved in or related to the processing of personal data of Vietnamese citizens and certain other people of Vietnamese origin. Decree 330’s key provisions cover the following areas: Administrative penalties for violations relating to the protection of national security and public order in cyberspace, including the dissemination of unlawful, false, or unverified information. Sanctions for cyberattacks, unauthorized access, introduction of harmful code or programs, and failure to cooperate with specialized cybersecurity forces. Sanctions for personal data protection violations, such as consent, cross-border data transfers, impact assessments, breach notification, and data-subject rights, among others—with maximum fines of up to 5% of an organization’s preceding-year revenue for cross-border transfer violations, or up to VND 3 billion for other data-protection breaches. Personal Data Protection Penalties The key sanctions for personal data protection violations are as follows: Consent violations: Fines of up to VND 70 million (approx. USD 2,642), plus potential additional sanctions and remedial measures including irreversible deletion of personal data collected without consent and confiscation of
Read More
August 25, 2026
Thailand’s Electronic Transactions Development Agency (ETDA) is studying potential new regulatory measures for digital platform services that could significantly expand the country’s digital platform governance framework. The ETDA has already conducted one public consultation session on the proposed measures and will hold additional sessions on August 25 and September 2, 2026, covering five types of platform services under the Royal Decree on Digital Platform Services B.E. 2565 (2022). The measures under study are preliminary and may be changed based on consultation outcomes. Foundational Measures Applicable to All Platform Types Seven baseline obligations would apply across all digital platform categories: Transparency reports. Platforms must prepare and publish statistical reports on platform governance activities, including the number of content items removed or restricted and appeal outcomes, in a comparable format. Notice and action mechanism. Platforms must establish minimum standards for channels to report potentially illegal content or goods, conduct case-by-case review, provide explanations when content is removed or restricted, and maintain an internal appeals channel. Rights over automated decision-making. Users significantly affected by automated decisions are granted rights to request an explanation, request human review, and contest the decision. Service level agreements (SLAs). Platforms must publish minimum standards for response times, processing timelines, progress notifications, and remedies for incidents on the platform. Labeling of AI-generated content. Content generated or modified by AI must carry visible labels and machine-readable metadata, with exceptions for creative works that disclose AI use in a nonmisleading manner. Prohibition of dark patterns. User interface designs that deceive, coerce, or distort user decision-making are prohibited, including hiding critical information, creating false urgency, or making service cancellation unreasonably difficult. Business user fairness. Platforms must meet minimum standards for the treatment of sellers, workers, and content creators, including advance notice of term changes, explanation of account suspensions or visibility reductions,
Read More
August 20, 2026
Thailand has established a new cross-ministerial committee to oversee data center operations nationwide. On August 5, 2026, the Thai cabinet approved the Prime Minister’s Office Regulation on the Data Center Business Policy Committee, which was published in the Government Gazette on August 13, 2026, and is now in effect. The regulation reflects the government’s policy to elevate Thailand’s digital economy and promote investment in digital infrastructure and AI. The key features of the new committee are outlined below. Definition of “Data Center” Under the regulation, “data center” is defined as a building, premises, or structure that uses electronic equipment to provide services related to the collection, storage, processing, hosting, or transmission of data by electronic means to third parties that are not affiliates, as further determined by the Data Center Business Policy Committee. Committee Composition The committee will be chaired by a deputy prime minister designated by the prime minister, and will have three vice-chairs comprising the ministers of digital economy and society, interior, and energy. The committee also includes 12 ex-officio members: the permanent secretaries of finance, agriculture, natural resources, energy, interior, digital economy, industry, and commerce; the secretaries-general of the Board of Investment (BOI), Energy Regulatory Commission, National Broadcasting and Telecommunications Commission (NBTC), and National Water Resources Office; and the director of the Energy Policy and Planning Office. Up to three expert members may be appointed by the prime minister for two-year terms, renewable once. The secretary-general of the National Economic and Social Development Council (NESDC) serves as member and secretary, with up to two NESDC officials serving as assistant secretaries. Powers and Duties The committee is empowered to: Propose policies, standards, and operational frameworks for government agencies in approving, licensing, issuing investment promotion certificates, or providing services to data center operators in Thailand; Study, analyze, and
Read More
August 14, 2026
Thailand’s Office of the Insurance Commission (OIC) has issued guidelines clarifying the boundaries between permissible and prohibited activities for unlicensed individuals—including influencers, bloggers, and content creators—when communicating about insurance products on social media. The Good Practice Guidelines for Persons Not Licensed as Insurance Agents or Brokers Regarding the Dissemination of Insurance Content Through Digital Media B.E. 2569 (2026) took effect on July 24, 2026. Activities Requiring a License The guidelines reserve the following activities for licensed agents and brokers: Soliciting or facilitating insurance contracts. Providing personalized advice on product suitability. Recommending policy cancellation to purchase promoted products. Creating links that facilitate contract formation. Receiving performance-based compensation tied to policies or premiums generated. Importantly, boilerplate disclaimers such as “this is not a recommendation to buy insurance” will not shield individuals from liability if the OIC views the content as personalized advice or solicitation. Permitted Activities Unlicensed persons may present general educational content about insurance—such as explaining terminology, sharing industry statistics, reporting news, or sharing personal experiences—provided the content does not target specific individuals to purchase from specific companies. The guidelines also set out best practices for communication, including presenting information in a fair and balanced manner that covers both benefits and limitations, encouraging consumers to read policy terms and consult licensed professionals, verifying information from credible sources before dissemination, and exercising special care when the audience may include vulnerable groups such as persons aged 60 and older. Prohibited Practices Prohibited practices include fear-based marketing, creating artificial urgency, omitting material limitations, making exaggerated claims, falsely claiming professional credentials, using fake engagement mechanisms, and sharing false or misleading content. The guidelines also reinforce the prohibitions under section 83 of the Life Insurance Act B.E. 2535 and section 78 of the Non-Life Insurance Act B.E. 2535 against soliciting insurance contracts with foreign operators
Read More