AI & Automation Trends: 2024 Insights & 2025 Outlook

As we reflect on the transformative developments in AI and automation throughout 2024, several key trends have emerged that are shaping the future of information management. During a recent webinar, AIIM Florida Chapter Chairman Craig Laue and I discussed the top industry news and research of the year and what they mean for the future of the industry. Based on that discussion, here's what we've learned and what we can expect in 2025.

Digital Transformation: The Paper Challenge Persists

Despite the buzz around digital transformation a few years ago, a surprising reality remains: over 45% of business processes are still paper-based, with some sectors showing even higher percentages. As noted in Deep Analysis's findings from the Market Momentum Index: In his blog post AIIM Board Member Alan Pelz-Sharpe explains that "those endless mountains of paper won't go away by themselves, nor will a wave of an AI wand make them magically disappear."

Where organizations used to retain paper documents for compliance or archival purposes alone, more are realizing there is potential value in digitizing paper documents to build datasets for AI.

The challenge of managing paper, digital, and AI-driven processes continues to make the role of information management practitioners increasingly complex, but they are well-positioned to help organizations scale their paper mountains. 

The Automation Maturity Gap

Recent findings from AIIM's State of the Intelligent Information Management Industry Report revealed a startling lack of automation maturity at the enterprise level. The numbers tell a compelling story:

• Only 33% of respondents reported having integrated systems or workflow and process automation in their team or department
• A mere 3% of respondents reported their team or department having attained workflow and process automation where advanced automation via Robotic Process Automation (RPA), and Artificial Intelligence/Machine Learning (AI/ML) technologies.

As organizations continue to invest in AI through 2025, we expect to see AI enhancing automation by adding intelligence and decision-making capabilities. However, a critical prerequisite remains: processes must first be properly documented and structured.

The AI Readiness Paradox

A clear dichotomy has emerged between AI development and organizational readiness. While 77.4% of respondents in the AIIM Market Momentum Index were either experimenting or in production with AI, significant barriers to success persist.

AIIM's State of the Intelligent Information Management Industry Report found that the majority of respondents (77%) rated their organizational data as either average, poor, or very poor in terms of quality and readiness for AI. This finding aligns with AvePoint's AI and Information Management Report 2024, which revealed that although 80% of organizations believed their data was AI-ready, nearly every organization surveyed (95%) faced data challenges during AI implementation, with over half (52%) encountering issues related to internal data quality and organization.

Based on conversations with our members throughout 2024, a pragmatic approach to data quality has emerged: focus on scope management rather than attempting wholesale organizational change. Instead of pursuing the potentially unattainable goal of organization-wide data quality, success lies in targeting specific datasets tied to AI experiments. By starting with discrete, manageable data hygiene projects, organizations can develop replicable processes and methodologies. These focused efforts create a blueprint that can be scaled to other datasets, building data quality capabilities incrementally rather than attempting a massive, one-time transformation. This iterative approach not only delivers immediate value for AI initiatives but also establishes sustainable practices for long-term data management.

Employee Engagement is Essential to AI Success

Success with AI implementation hinges on cultural acceptance and employee engagement. AIIM's Organizational Readiness for GenAI highlighted three crucial pillars:

1. Content Access
2. Content Quality
3. Employee Engagement

Author Rob Bogue explained in the white paper that "employees must be supported in a way that ensures their safety while exploring opportunities for benefit."

"Large-scale investments in modern technology fail not because of technical limitations or challenges, but because users refuse or lack the training to use the technology in ways that enhance the organization's performance," wrote Rob Bogue. This is supported by Market Momentum Index findings, where 22% cite users and stakeholder adoption as a key obstacle to effective leveraging of AI, while 33% said lack of skilled personnel was an obstacle.

The Evolution of AI: RAG and Agentic AI

Advancements in AI have altered the way organizations contextualize, find, and use unstructured data. 

Retrieval-Augmented Generation (RAG)

Retrieval-Augmented Generation (RAG) has emerged as a crucial development in AI implementation. As McKinsey & Company explains: "RAG allows LLMs to access and reference information outside the LLMs own training data, such as an organization's specific knowledge base, before generating a response—and, crucially, with citations included." RAG has improved the reliability and security of generative AI in the enterprise. 

While RAG technology offers powerful capabilities to leverage organizational knowledge, its effectiveness ultimately depends on the quality of underlying data sources. Throughout 2024, conversations with CIOs and information leaders have revealed a common challenge: ambitious RAG implementation initiatives often stumble when confronted with disorganized or poorly maintained file storage systems, including platforms like SharePoint. This reality check has highlighted that even advanced AI technologies require a strong foundation of well-structured, high-quality data.

As we move into 2025, information leaders will play a crucial role in bridging this gap, helping enterprises prepare their data infrastructure for successful RAG implementations. Microsoft Copilot stands as a prime example of RAG's potential in the enterprise space, but its success stories often share a common thread: organizations that invested in data quality before deployment saw the best results.

Agentic AI

Looking ahead, Agentic AI represents the next frontier.

According to IBM: "Agentic AI refers to a system or program that is capable of autonomously performing tasks on behalf of a user or another system by designing its workflow and using available tools." While still an emerging technology, Agentic AI builds upon generative AI models to create more autonomous and capable systems.

A key benefit of agentic AI is it's ability to adapt and learn from unexpected data. It's fair to say that the majority of unstructured data within an organization can be messy and even unexpected. Different from robotic process automation, agentic AI promises to process unstructured data, apply context, and make decisions. Yet, as this helpful article from CIO explains, the engineering effort behind agentic AI is significant. 

I predict that agentic AI is still in very early experimental stages and the required technology infrastructure, engineering, and skillsets needed for successful implementation will slow adoption for most organizations. However, information leaders can serve as valuable partners in agentic AI implementation, helping project teams safely expose the appropriate data to ensure the agent has the needed context to complete a given task.  

Looking Ahead: The Convergence of Information Management and AI

As we've seen throughout 2024, the intersection of information management and artificial intelligence presents both significant opportunities and complex challenges. The path forward isn't about replacing traditional information management practices, but rather evolving them to meet the demands of the AI era. Whether organizations are still scaling their paper mountains, implementing RAG solutions, or exploring the frontier of Agentic AI, one truth remains constant: the quality and accessibility of information/unstructured data remains fundamental to success.

Join us at the AI+IM Global Summit (March 31-April 2, 2025, in Atlanta, Georgia) to dive deeper into these challenges and opportunities. Through curated interactive workshops and focused sessions, you'll gain practical strategies for leveraging unstructured data and implementing AI solutions that deliver real value. Don't miss this opportunity to connect with peers and experts who are successfully navigating the convergence of information management and AI. Learn more and register.