Online: Trending Now

We have entered an era of greater scrutiny and fewer resources in higher education. How might AI enable the deans to become more effective and efficient in administering the college?

The challenges are proliferating while funding is deteriorating. Fortunately, the AI options to accomplish more with less funding are expanding. As of the end of February, a number of awe-inspiring “deep research” tools have been released. More than half a dozen such tools are available from different providers at prices ranging from no cost to $200 a month. They are becoming the key to enhancing efficiency and effectiveness of administrators, including deans.  Omar Santos, distinguished Cisco engineer, published on February 21, 2025 “A Comparison of Deep Research AI Agents” where he outlines some of the features of the five leading brands as of that date, noting, that “unlike simple question-answering bots, these agents perform multi-step reasoning: formulating search queries, browsing web content, analyzing data, and synthesizing findings into structured outputs with citations.” Santos goes on to describe that there are two primary architectural approaches to deep research agents:

Fully Autonomous Agents: Once given a prompt or topic, these agents operate independently end-to-end. For example, OpenAI’s Deep Research feature (launched in Feb 2025) allows ChatGPT to act like a “research analyst”, working for several minutes without intervention to gather information from the web and compile a report with sources. It is powered by a specialized version of OpenAI’s upcoming o3 model optimized for reasoning and web browsing. The user simply provides the topic, answers a few additional questions, and the agent handles the rest autonomously. This fully-automated approach is convenient but requires a very robust agent to decide on research directions and verify information on its own.

Human-in-the-Loop (HITL) Agents: These agents incorporate human feedback or approval at different steps of the research workflow. Rather than running to completion unquestioned, the AI will pause for guidance — typically after formulating a research plan or outline — so the user can review and adjust it before the agent proceeds.

Santos compares and contrasts five different such tools, leading with the OpenAI and Google Gemini versions. I would add two more to the list. First is the outstanding “Storm” tool developed as a brainstorming device by the Open Virtual Assistant Lab (OVAL) at Stanford University. The other is the Grok 3 Beta recently released by X.ai. In all cases, these tools are capable of using advanced reasoning to develop research plan; searching the internet at large and other selected sites to which you provide access permissions; conducting probing research; composing a report with citations; revising the report and updating as directed.  Increasingly, the tools are offering options to ensure they do not use your inputs for training. Each of these tools will, no doubt, revise and improve in the coming months as new competitors enter the field. I expect that university IT departments will assist deans and other administrators as they select, train and become proficient at using the tool best suited to their needs.

Here are ways these tools may assist deans in meeting the challenges of their positions this fall and moving forward.

Personal Assistant

• These very “smart” tools can manage calendars, set reminders, respond to routine correspondence and more.  In these cases, initially deans may want to give individual approval of actions, but in time, just with an experienced assistant, they may want to enable auto-processing while keeping copies for follow-ups.
• One can share an email or notes of a conversation with any additional points that should be included in the response and ask the deep research tool to compose a response (including references it discovers that are relevant to the communication).
• The deep research tools can automatically schedule meetings and prepare agendas for items that either the dean or the tool may identify as emerging issues for the college.

College Research projects

• A continuing assignment may be to conduct a weekly search for new public and private funding of projects in which the faculty of the college have an interest.
• Strategies for reducing indirect costs of projects can be researched and a report shared with project managers, department chairs and faculty conducting research.
• Focused reports can be generated to propose extended funded research topics and opportunities in areas where the college faculty have conducted preliminary research.
• Emerging markets for products of research can be identified and letters of introduction to businesses who might value the research can be drafted and sent.

Curriculum Currency and Relevancy

• The Deep Research tool can compare the college’s published curriculum with those of peer institutions and others for timeliness, utility, and corporate demand.
• A dynamic comparison of the top ten competitor colleges’ curriculum and research agendas can be maintained with update alerts when a competitor makes changes.  In such cases, the tool can automatically create a meeting of relevant faculty and staff, including an agenda with materials from the competition to focus the discussion.
• Deep Research can conduct predictive analyses of current curriculum, identifying courses that obstruct the smooth flow of students through the curriculum (once identified, the tool can set up an agenda with data handouts to discuss the problem and suggest solutions). In such studies learning outcome effectiveness can be assessed, percentage and time to employment of graduates or certificate completers, and retention of graduates by employers can be analyzed.

Meeting and Leading the Competition

• Deep Research is the ideal tool to identify new domestic and international markets for enrollments.
• Professional certificates can be designed by the tool to meet emerging needs in the field. These can be matched to faculty backgrounds for potential staffing.
• Deep Research can do an analysis of faculty workloads; identify those who may need more support and those may be available for more activities; and make recommendations to the dean and department chairs.

Analyzing and Assessing Productivity and Currency of Unit Work

• Deep Research tools are able to extrapolate on the work of current projects and compare their objectives to emerging markets, technologies, and societal needs. Sharing such reports with the relevant units as well as preparing the agenda for discussion, keeping minutes of the meeting, and codifying outcomes can all be accomplished with AI tools.
• Full annual reviews and analysis of revenue generated, students enrolled, outcomes accomplished, and other such data can be accomplished by Deep Research. These can help to guide strategic planning.

These are just a few of the important tasks of the dean that can be assisted by Deep Research tools. There are many more tasks that can be tackled by these tools. I hope that this brief list will prompt readers to become comfortable with the range of work that can be done in order to identify their own tasks for which they could use assistance.

Is your university preparing to implement these tools in support of deans and other administrators? Has training begun? It is important that your institution gets started so that you will not rapidly fall behind your peers in utilizing advanced Ai tools to enhance effectiveness and efficiency.

 

This article was originally published in Inside Higher Ed.