In the previous post in this series, I explored what happens when AI-powered systems challenge the values of libraries, archives, museums, and knowledge organizations. Those values can help institutions decide where AI belongs, what safeguards are required, and where limits should be placed. The next question is: who gets to put those principles into practice?
Discussion about the future of information work often concentrates on tasks: which activities AI might perform, which skills professionals will need, and which roles may change. Yet the future of the profession will also depend on who selects the systems, shapes the workflows, evaluates the results, and develops the expertise required for the work ahead.
Drawing on recent research and examples from professional practice, I want to consider what remains constant as technology changes, how professional judgment is developed, and whether new responsibilities will bring meaningful influence.
As AI becomes more widely embedded in organizations, will information professionals help shape the systems through which knowledge is found and interpreted—or be expected to correct problems after the important decisions have been made?
An Important Consideration: What AI Does Not Change
A recent systematic review of the evolving roles of information professionals identifies a tension between empowerment and deprofessionalization.
Automation may reduce repetitive work and free up information professionals to take on more analytical or strategic responsibilities. It could also diminish professional autonomy if systems and workflows are determined elsewhere. Released capacity may support better services, but it can just as easily disappear into reduced staffing or an expanding workload.
However organizations redistribute individual tasks, the fundamental purpose of information work remains remarkably consistent: helping the right person find the right information when it is needed.
My own email archive offers a small example. Older LIS-LAW messages include requests for missing judgments, obscure books, and unavailable publications. When databases and catalogs failed, information professionals turned to colleagues who might know where an item was held or have access to a copy.
AI may accelerate the initial research, but it cannot provide a verifiable substitute for a source it cannot access. If a lawyer needs an obscure unreported judgment, a generated account will not do. The document itself remains the evidence and must still be located and verified.
This is where the apparently old-fashioned professional network retains its value. The technology used to find the connection may change, but the responsibility remains: distinguish an answer from a lead and recognize when an inquiry remains unresolved. As generated answers become easier to produce, knowing whether the underlying evidence has actually been found becomes increasingly important.
Reclaiming the “Middle Space” Between People and Technology
During a Syracuse University iSchool panel on the AI moment in libraries, Sanda Erdelez observed that information systems increasingly generate material rather than simply retrieve it. She argued that librarians must reclaim the “middle space” between people and technology.
Information professionals have long occupied that space. They connect users with sources, clarify questions, interpret results, and notice what may be missing. Their decisions about collections, metadata, taxonomies, access, and preservation also shape discovery behind the scenes.
Generative systems change this relationship. A user can receive a fluent synthesis without encountering the underlying collection or research process. The system appears to select the material and supply its meaning.
Job advertisements preserved in my email archive illustrate how professional boundaries had already been moving. Roles once centered on library administration and information retrieval expanded to include SharePoint, internal websites, content management, systems, data governance, and consultancy. More recent vacancies and professional discussions refer to AI readiness, knowledge transfer, persona-driven services, and “knowledge as a product.”
My archive cannot establish a profession-wide trend, but it shows how changing expectations appeared in ordinary material circulating through one career. Responsibilities expanded while the underlying concern with organizing knowledge and making it usable remained recognizable.
Erdelez describes the emerging response as “curating algorithms.” In practice, information professionals may influence the knowledge available to AI systems, determine which sources should be treated as authoritative, test how information is represented, and decide where human review remains essential.
New Responsibility Does Not Guarantee New Authority
The distinction between responsibility and authority is central to the future of information work.
A librarian may be expected to identify errors in AI-generated metadata without being able to alter the system that produced it. A knowledge manager may be held responsible for the reliability of an enterprise assistant while having little influence over the repositories it searches. A research specialist may teach users to question generated answers after a vendor has embedded AI into a subscription platform.
In each case, the professional remains responsible for information quality while working downstream from the decisions that determine it.
Meaningful influence begins earlier. It includes participation in procurement, source selection, workflow design, evaluation, and decisions about acceptable risk. It also requires the ability to challenge or constrain unsuitable applications. Otherwise, keeping a “human in the loop” may assign responsibility for failures to someone who had little power to prevent them.
How Will the Next Generation of Information Workers Develop Judgment?
If AI assumes more routine searching, summarizing, metadata creation, and administrative work, how will newcomers develop the judgment needed to evaluate its output?
Routine work can carry considerable developmental value. New professionals learn through ambiguous inquiries, incomplete records, and unexpected exceptions. They begin to recognize failure by observing the research process rather than seeing only its final result.
I saw this when I interviewed Laura about her graduate traineeship at the Institute of Advanced Legal Studies. Her experience included inquiry work, research and conservation training, participation in a library-system change, and contact with other institutions. She developed confidence in an environment where she could ask questions and learn from experienced colleagues.
Her experience does not prove that AI will damage professional development. It shows how judgment is built. If automation removes parts of that pathway, organizations will need to create other opportunities for supported practice, reflection, and increasing responsibility.
Future Readiness Is an Organizational Capability
One practical framework for examining these wider conditions is APQC’s Knowledge Management Capability Assessment Tool. It assesses KM maturity across four connected areas: strategy, people, process, and content management/IT.
These include objectives and funding, governance and leadership, knowledge flows, measurement, content management, and the supporting technology.
Applied to AI, the framework moves the discussion beyond individual skills. Organizations can ask whether an AI initiative addresses a genuine knowledge need, whether the right people can influence its design, how information will be reviewed, and whether the underlying content is reliable enough for the proposed use.
Future readiness depends on the organization’s ability to connect its people, knowledge, processes, and technology—and to recognize where professional judgment belongs within that system.
The Future Role Is Up to Us
Information professionals have long worked at the center of a web of people, sources, systems, and institutional knowledge. As we have seen, the tools and responsibilities surrounding that work continue to change. The need to locate reliable evidence, recognize what is missing, and bring the right knowledge together remains.
AI adds a new layer to that web. It can generate answers and recommend action, but people still determine which sources it can reach, how its output is assessed, and where its authority ends. Information professionals have the expertise to help make those decisions, provided they are involved early enough to influence them.
The answer to the title therefore rests with organizations as much as with the profession. New skills will matter, but so will job design, professional development, and decision-making authority. As AI becomes more widely embedded, will information professionals simply work with the systems they are given—or help shape how those systems work?
0 Comments