Artificial intelligence (AI) is transforming the way businesses operate. Employees at many companies routinely use AI tools to complete tasks more efficiently. The same is true for employees at their customers, vendors, and partner organizations.
Businesses whose value depends on proprietary content – such as research reports, training programs, methodologies, or other similar intellectual property (IP) – face unprecedented challenges in protecting it in the age of AI. These challenges apply not only to how the company uses AI internally. A bigger issue is what happens when external parties have access to powerful AI tools that can analyze, summarize, replicate, and share proprietary content in seconds.
Content Leaving the Building in Multiple Ways
Most companies are aware that employee misuse of AI tools poses certain risks. Employees may, for instance, paste the proprietary content of their employer into a public AI tool, inadvertently exposing the proprietary content to wider integration within a large language model (LLM).
But internal misuse of AI tools is only part of the challenge. Businesses who make their proprietary content available to customers, licensees, and other third parties must adapt to the reality that AI provides these external parties with unprecedented opportunities to misuse the content.
Replication Risks
Third parties can now use AI tools to efficiently create substitutions for proprietary content – eliminating the need to continue paying for the product or service they purchased to get access to the proprietary content. For instance, if a company produces educational materials, AI tools can be employed to summarize and reorganize the information, create facilitator guides and student workbooks, and build a similar program – allowing customers to become former customers and even competitors in some instances.
Many existing licensing agreements were written before AI tools made large-scale content replication feasible. As a result, standard IP and use-restriction provisions do not adequately guard against modern risks.
Some content-driven companies have responded to the proliferation of AI use by inserting blanket "no AI" clauses with respect to their proprietary materials into their customer contracts. But these strict clauses are extremely difficult to monitor and enforce in practice, and are often met with significant customer pushback, because AI tools have become embedded in standard business workflows.
While AI prohibitions may still be advisable in certain situations, a more nuanced and effective approach is to laser-focus IP contract language on anticipated user behavior, and ensure that all users are educated on the restrictions. For example, a modern AI clause restricts derivative works AND requires that all users be educated on the restrictions. I have started to see liquidated damages provisions tied to breach of AI-use restrictions. While these clauses are tough to negotiate, such language is worth considering for certain kinds of proprietary content. When language restricts what the customer can produce rather than what tools they can use, contract provisions are easier to sell and to enforce.
Creating Other Barriers
When content is delivered as downloadable PDFs, workbooks, or PowerPoint presentations, it can be uploaded into AI tools in seconds. However, when proprietary content is delivered through a carefully calibrated and controlled platform –such as a membership website, subscription portal, or learning management system (LMS), a meaningful barrier exists that reduces large-scale extraction of the content. Platforms can be configured to limit or prohibit downloading content, and the user's interaction with content can be tracked to identify potential license breaches.
Another practical deterrent is to embed a detailed, explicit copyright and use-restriction notice as a footer on every page of content, rather than simply including the copyright symbol or placing a restrictive notice on the cover page. Most AI tools now contain guardrails designed to avoid reproducing or altering material that carries strong copyright indicators. Ubiquitous use of restriction notices is more likely to trigger these guardrails.
AI-Assisted Content and Copyright Protection
In addition to monitoring how others use their proprietary content, companies need to control internal AI practices in order to ensure their content is eligible for protection in the first place. With organizations increasingly relying on AI to help create a wide range of IP, there is significant risk as to whether the resulting work is protected.
Content generated entirely by AI is not eligible for copyright protection. However, the use of AI as a tool does not eliminate protection, as long as a human author shapes the final expressive content.
To protect their content and help combat downstream replication, companies need to institute and enforce a strict internal AI policy for content creation. While AI can be used to create first drafts, meaningful human input such as editing, selecting which information to add or remove, and other creative contributions must be made for protections to be viable.
Looking Ahead
To protect their interests in the age of AI, companies whose value depends on proprietary content must be proactive. Companies need a robust AI policy that governs what content goes on the page, how it is delivered, and how it can be used once it leaves their premises. Companies who use AI tools to create IP must make significant human creative contributions and document those contributions. Customer contracts should be updated to include targeted restrictions on replication and derivative use of proprietary content. Companies can also create meaningful barriers to replication by delivering content through a controlled platform and by placing copyright and restriction notices on every page. AI is not going away, and companies need to adapt their content protection strategies to the new realities of the marketplace.
Mendi Sossamon, a Partner at Potomac Law (PLG), crafts legal solutions to help businesses grow. She leverages more than 27 years of legal experience, which includes a stint in Big Law, running her own law firm and serving as a Deputy General Counsel for a billion-dollar company. Based in Austin, Texas, Mendi focuses her practice on outside general counsel services, complex corporate and commercial transactions, intellectual property strategy, and commercial optimization.
