Content Licensing Do’s and Don’ts in the Age of AI

  October 16, 2024  |   

Artificial Intelligence’s (AI) seemingly endless capabilities and applications present great opportunities (and some challenges too) for publishers and societies across the publishing enterprise. One of the main areas of both growth and reason for caution to emerge is the potential to license scholarly content to AI providers—primarily to be used to “train” large language models (LLMs).  While this type of licensing opportunity may be compelling, it requires thoughtful integration into the organization’s overall content portfolio management and revenue strategy.

Recently announced licensing agreements between scholarly and academic publishers and technology companies highlight AI’s insatiable demand for primary, verified, reliable information. AI developers rely on this high-quality, vetted content to train models, refine algorithms, and enhance natural language processing capabilities. This demand can present a lucrative opportunity for publishers to license content – aka the knowledge needed for training. It also raises important strategic questions about ownership, sustainability, and long-term business models that should not be ignored in the process.

Opportunity vs. Risk: Licensing Content Do’s and Don’ts 

If a partnership with an AI company seems intriguing, it is…as long as you proceed with an understanding of how this opportunity may play out for your organization and where on the classic innovation adoption curve you are comfortable. Here is a handy checklist to help you evaluate the opportunities and risks of licensing content to AI providers. Keep in mind, YMMV, as will your priorities.

Do:

  • Integrate Licensing into Overall Content Strategy – View AI licensing as part of a broader content portfolio management plan to align with business objectives and sustain long-term value.
  • Prioritize Content Based on Value – Categorize content by demand and monetization potential to tailor licensing strategies for different segments (e.g., niche vs. broad appeal).
  • Introduce Strategic Pricing Models – Experiment with flexible pricing strategies like volume-based, usage-based, or hybrid models to reflect content value and accommodate AI providers’ diverse needs.
  • Complement and Enhance Existing Revenue Streams – Ensure that AI licensing supports rather than undermines other revenue channels (subscriptions, APCs, institutional licensing, etc.). Consider tiered access or differentiated pricing for recent vs. older content.
  • Collaborate with AI Companies Ethically – Build partnerships that ensure responsible content usage. Establish guidelines for ethical AI content generation, labeling, and attribution.
  • Protect Author Rights – Ensure that licensing agreements comply with existing contracts and protect authors’ rights. Proactively manage relationships with scholars to maintain trust and uphold their interests.
  • Be Prepared for Market Shifts – Experimentation is the order of the day but the market and innovation is moving fast.  Adopt flexible frameworks to quickly adjust to technological changes or shifts in demand for licensed content.
  • Maintain Transparency and Communication – Keep authors, research communities, and internal stakeholders informed about how the organization’s content is licensed and used by AI firms.
  • Consider Partnering with Other Content Providers – Strategically partner with publishing peers to offer a broader range of niche content. Collectively negotiate through a ‘power in numbers’ approach.

Don’t:

  • Rely Solely on AI-Driven Revenue – Avoid becoming over-reliant on revenue from AI licensing, as market shifts could jeopardize financial stability if demand for licensed content declines.
  • Undermine Content Value – Be cautious of pricing models that risk devaluing content over time, especially as AI-generated content becomes more sophisticated.
  • Ignore Unintended Consequences – Don’t overlook the potential for content devaluation or the blurring of lines between original research and AI-generated outputs.
  • Neglect Author Concerns – Don’t disregard the potential for author questions, dissatisfaction, or misuse of their work. Always respect contractual obligations and maintain productive relationships with the academic community.
  • Overlook Ethical Concerns – Avoid participating in licensing agreements without ensuring ethical guidelines for the use of AI-generated content, including issues like data privacy and security.
  • Ignore the Long-Term Impact on Scholarly Publishing – Don’t assume AI-driven licensing won’t affect traditional publication models. Proactively assess how AI might impact and change peer review, publication demand, and researcher incentives.

Final Thoughts

Licensing content to AI providers is certainly a potential opportunity for publishers. That opportunity also comes with possible risks and the need for some caution. These Do’s and Don’ts serve as a starting point to help you begin to frame out how partnerships with AI providers may or may not “fit” with your strategy, mission, and organizational goals, while acknowledging the need to consider safeguards to protect the integrity of your content, author relationships, and long-term sustainability.

Delta Think can help your organization understand the unique opportunities and challenges of integrating AI licensing into a comprehensive content portfolio management strategy.  Ready to start the conversation? Contact us today.

As Ideas in Action went to press, Ithaka S&R announced a Generative AI Licensing Agreement Tracker to help capture the details, impact, and strategy of these deals.