Generative artificial intelligence has quickly become a routine business tool. Employees use AI platforms to draft marketing materials, prepare reports, create presentations, write software code, develop product descriptions, and generate artwork. Independent contractors are increasingly relying on the same technologies to perform services for their clients. As the use of these systems expands, businesses are beginning to ask an important question: Who owns the content generated by artificial intelligence?
A Paradox
Traditionally, intellectual property ownership has been relatively straightforward. An employee creating work within the scope of employment generally produces work owned by the employer. Independent contractors may own their work unless an agreement provides otherwise. The use of generative AI slashes costs and increases productivity, but completely scrambles the traditional rulebook on copyright ownership.
Under current U.S. law, copyright generally protects original works of authorship created by human beings. The U.S. Copyright Office has repeatedly stated that content generated solely by AI, without meaningful human creative input, is not eligible for copyright protection. If a user enters a brief prompt and an AI platform produces an article, image, or other content, copyright protection is unavailable for the work produced.
This does not mean that all AI-assisted content lacks protection. Many projects involve substantial human contributions. A user may develop detailed prompts, select among multiple outputs, edit the material extensively, rearrange content, and add original expression. In such situations, the human-created portions of the work may qualify for copyright protection even if certain AI-generated elements do not.
AI Platforms
Another issue concerns the contractual terms governing AI platforms. Many users focus on what the technology can do without carefully reviewing the applicable terms of service. Different AI providers may allocate rights differently. Some AI providers state that users own the output generated via the AI platform, while other AI providers reserve certain rights to use, store, analyze, or improve their systems using submitted content. Businesses that rely heavily on AI-generated materials should understand the contractual rights associated with the AI platforms they use.
Employee Use of AI
Ownership questions become even more complicated when employees are involved. Suppose an employee uses an approved AI platform to create marketing materials for the company. The employer may assume that the resulting content belongs to the business. While that may often be true as a practical matter, companies should not rely solely on such assumptions. Employment agreements should clearly state that all work product created during employment, including work generated with the assistance of artificial intelligence tools, belongs to the employer.
Independent Contractor Use of AI
Independent contractors present a similar challenge. Many businesses hire consultants, graphic designers, software developers, and marketing professionals who routinely use AI tools in providing their services. Unless a written agreement addresses ownership, disputes can arise regarding both the human-created and AI-assisted portions of any deliverables created in the course of such work. A business’s service agreements should expressly provide that all intellectual property rights associated with any work product are assigned to that business, to the fullest extent permitted by law.
Human Supervision
Businesses should also consider the possibility that AI-generated content may incorporate elements resembling existing copyrighted works or trademarks. Although AI providers attempt to reduce these risks, concerns remain regarding the origin of training data and the possibility of the creation of infringing outputs. Human review remains essential before publishing or commercializing AI-generated content.
A practical solution is to adopt clear policies and contracts. Employment agreements, independent contractor agreements, confidentiality agreements, and intellectual property assignment provisions should all be reviewed in light of evolving AI technologies. Companies should identify approved AI platforms, establish rules governing their use, and clarify ownership expectations – before disputes arise.
AI has the potential to constitute a modern gold rush for businesses, promising unprecedented productivity and slashed costs. But as it stands, this new frontier also presents a wild west for content and copyright ownership issues. We urgently need Congress to pin on the sheriff’s badge and craft some definitive legal parameters for copyright and ownership when it comes to generative AI work product. Until then…Be Careful!
