Accessibility Meets Engagement: From Compliance to Competitive Advantage
The accessibility discussion drew standing-room-only attendance and sustained energy throughout. The shared discomfort around accessibility challenges created space for honest conversation about what many described as a daunting and expensive undertaking, particularly for legacy archives.Yet the session consistently reframed accessibility as opportunity rather than obligation. Participants emphasized that accessibility represents a core component of open science rather than a compliance checkbox. The real return on investment comes from ensuring the entire research community can engage with published scholarship, not simply meeting WCAG or ADA thresholds.
The conversation surfaced strong consensus around moving accessibility upstream into author, editorial, and platform workflows rather than treating it as a post-publication remediation challenge. While automated tools can accelerate progress, human oversight remains essential. For massive archival collections, participants acknowledged that "good enough to start" may become a practical necessity, with AI-powered remediation at scale potentially offering a path forward even if it requires accepting some quality trade-offs.
Publishers are also navigating increased transparency expectations from libraries, which now request multiple forms of documentation beyond VPATs. Being open about roadmaps, progress, and feedback loops emerged as critical to building trust with institutional partners. Perhaps most importantly, the group explored how to shift the narrative from regulatory pressure to innovation incentives—moving from "sticks" to "carrots" that reward accessibility as a driver of discoverability, engagement, and mission alignment.
AI as Editorial Partner: Strategic Pathways for Workflow Integration
The AI integration discussion took a structured approach to evaluating implementation strategies. Participants examined three distinct pathways: Professional Evolution (AI-enhanced human decision-making), Industrial Automation (AI-first processing with human oversight), and Community-Focused Evolution (discipline-specific AI tools). Each pathway offers different approaches to addressing operational challenges and research integrity risks.Using a strategic risk-reward framework, tables identified "quick win" opportunities that could deliver near-term value versus longer-term investments requiring more substantial organizational change. The conversation explored critical build-versus-partner decisions, recognizing that different publishers bring different capabilities to AI implementation.
Participants discussed organizational readiness requirements, realistic pilot timelines, and how to establish meaningful success metrics. The discipline-specific nature of scholarly publishing means stakeholder acceptance varies significantly across different research communities. Publishers left with clearer understanding of which strategic pathway aligns with their operational strengths and concrete next steps for piloting AI solutions that could reduce processing costs, accelerate publication timelines, and strengthen research integrity protections.
Humans, Bots, and Discovery: The Traffic Attribution Challenge
The bot traffic and discovery session tackled one of scholarly publishing's newest operational challenges: distinguishing between beneficial automated access and resource-draining traffic. As generative AI platforms reshape how researchers discover and consume content, publishers face competing pressures around traffic management and content exposure.Participants explored the tension between traditional SEO strategies and emerging GEO (Generative Engine Optimization) approaches. While search engine visibility has driven discoverability for two decades, the rise of AI-powered research assistants requires rethinking how content gets surfaced and consumed. The conversation examined how to balance protecting platform resources from excessive bot crawling while ensuring content remains accessible to the AI tools that increasingly mediate researcher workflows.
The session highlighted fundamental shifts in user behavior as scholars adopt AI assistants for literature review, research synthesis, and knowledge discovery. Understanding these evolving patterns will be critical for publishers seeking to maintain relevance in researcher workflows while managing the technical and business implications of automated content access.
Monetizing Content in the AI Era: Beyond Traditional Licensing
The monetization discussion revealed sophisticated thinking about alternative revenue pathways beyond bulk licensing to foundational AI models. Smaller publishers in particular are exploring strategies that maintain greater control while creating new value streams.The Model Context Protocol (MCP) emerged as a particularly interesting approach, enabling real-time LLM access to publisher content without traditional licensing transactions. Participants also discussed leveraging collective licensing mechanisms through organizations like the Copyright Clearance Center to strengthen negotiating positions for publishers who lack individual leverage. The conversation identified multiple use cases beyond training data provision, including RAG implementations for specialized queries, content access for corporate AI tools, and AI-enhanced products like search assistants and research tools integrated into society subscriptions.
However, critical concerns temper the enthusiasm around AI licensing opportunities. Publishers emphasized the need for protection from liability, control over derivative works, and avoiding lock-in with single AI providers. Technical challenges around data retention create uncertainty—once content is ingested into AI systems, publishers question whether it can ever be fully extracted when licenses expire.
The "black box" pricing model employed by AI companies emerged as a significant frustration. Without transparency about valuation methodology or market precedents, publishers struggle to assess whether licensing offers represent fair value. This information asymmetry puts smaller publishers at considerable disadvantage in negotiations.
Participants stressed that any AI partnerships must align with publisher codes of ethics, include appropriate usage reporting, and feature pricing models that different market segments can afford. The goal is creating products that add genuine value rather than simply providing raw content access.
Looking Ahead
Across all four sessions, publishers demonstrated commitment to advancing their mission while adapting to technological and regulatory change. Whether addressing accessibility mandates, integrating AI capabilities, managing new traffic patterns, or exploring alternative business models, the scholarly publishing community is approaching transformation with both pragmatism and purpose.
Though the breakout rooms weren't recorded, you can explore the rest of the 2025 archive here.