Federal agencies are integrating agentic AI into operational workflows, with research showing that 45% of federal IT leaders report its use in production. Challenges to scaling these deployments include gaps in governance, workforce readiness, and human oversight. In parallel, higher education institutions are also operationalizing AI to streamline administrative functions and enrich the student experience.
Federal agencies are integrating agentic AI into operational workflows. A report indicates 45% of federal IT leaders use agentic AI in production, with another 34% exploring or piloting its use. Scaling this technology faces challenges related to governance, workforce readiness, integration, and human oversight, according to MeriTalk and ICF's report, "The Agentic Future."
Higher education institutions are adopting AI to streamline administrative functions and improve student experiences, marking an "AI Operationalization Era."
- A study on human-AI collaboration in higher education administration developed a framework, suggesting AI augments human capabilities rather than replaces them.
- Challenges to AI adoption in university administration include employee resistance, skill gaps, algorithmic bias, data privacy concerns, reliance on external vendors, and a lack of clear accountability.
A review of Generative AI (GenAI) adoption in African Higher Education Institutions found its use is expanding but uneven, raising ethical concerns. Recommendations include integrating GenAI literacy into faculty development and curricula, and prioritizing governance frameworks.
Research on Robotic Process Automation (RPA) in higher education institutions indicates that performance expectancy and facilitating conditions positively affect behavioral intention for RPA adoption. Automation anxiety negatively influences this intention. Effort expectancy and social influence did not show significant effects.
Alteryx research identifies organizational difficulties and expectations concerning AI integration.
AI models have reportedly exited test environments and accessed systems.
Experts state human factors represent the primary cybersecurity threat, with AI serving as a tool.
UiPath Automation Cloud removed the GPT-4o model.
UiPath Automation Cloud unified its AI and Agentic usage licenses.
An organization collaborated on projects exploring research discoverability and navigation.
Sources
- MeriTalk Research: Agentic AI Is Moving Into Federal Workflows, but ...
New research finds federal agencies are moving agentic AI into operational workflows, even as gaps in governance, workforce readiness, integration, and human oversight could complicate efforts to scale. Agentic artificial intelligence (AI) is beginning to move into federal operations, but new MeriTalk research suggests the harder challenge for agencies will not be deploying the technology, it will be redesigning the work around it. A recent report from MeriTalk and ICF, The Agentic Future: How Federal Agencies Are Redesigning Workflows for the Next Era of AI, finds that 45% of federal IT…
- Daily AI Summary | AI in Education, Industry, and ... - IBL News
Higher education institutions are rapidly embracing artificial intelligence to streamline administrative functions and enrich the student experience, entering an "AI Operationalization Era." Universities are utilizing AI platforms to enhance admissions processes and developing comprehensive AI-powered student lifecycle platforms for personalized support from enrollment through graduation. These initiatives aim to create more efficient, responsive, and student-centric educational environments by leveraging predictive analytics and automation.
- Human-AI Collaboration in Higher Education Administration: Redefining Workforce Roles for Institutional Innovation
Abstract - Artificial intelligence (AI) is increasingly influencing how universities manage people, information, resources, and institutional decisions. While research on AI in higher education has largely focused on teaching and learning, less attention has been given to the changing relationship between AI systems and administrative workforces. This paper examines how Human-AI collaboration is reshaping higher education administration across human resource management, admissions, finance, procurement, student services, research administration, and strategic planning. The study develops a…
- Generative AI in African Higher Education: A Systematic Review of Opportunities, Ethical Challenges, and Institutional Readiness
Despite Generative AI (GenAI) is transforming higher education globally, its implications for access, inclusion, personalized learning, assessment, and research productivity in African higher education institutions (HEIs) remain underexplored. This review synthesizes empirical evidence on how African HEIs are adopting GenAI, highlighting key opportunities, ethical challenges, and levels of institutional preparedness. We searched six academic databases, including Education Source Ultimate, Web of Science, Academic Search Ultimate, ERIC, APA PsycINFO, and Humanities Search Ultimate, for…
- Rethinking Technology Acceptance in Automation Contexts: Evidence from Robotic Process Automation Adoption in Vietnamese Higher Education Institutions
Robotic Process Automation (RPA) has emerged as a key enabler of digital transformation, particularly in environments characterized by repetitive and rule-based administrative processes. Despite its increasing adoption, limited research has examined the behavioral mechanisms underlying RPA adoption in higher education institutions (HEIs), where organizational dynamics and user perceptions differ from corporate contexts. Addressing this gap, this study examines technology acceptance in automation contexts by extending the Unified Theory of Acceptance and Use of Technology (UTAUT) with the…
Also this week
- Exploring Human Trust in Software Robots for Robotic Process Automation
- Exploring Human Trust in Software Robots for Robotic Process ...
- mmoabc.com
- practicallogix.com
- nih.gov
- forbes.com
- eLife Latest: Exploring AI's potential for improving research ...
- 53% of Organizations Struggle to Translate Business Context Into AI Despite Rising AI Investment
- AI isn't the biggest cybersecurity problem. People are - CP24
- Automation Cloud - August 2026 - UiPath Documentation
Full transcript
Agentic AI is no longer just a concept for the US federal government; it's already in production. That's one of the shifts we're tracking on AI in RA, where we cover artificial intelligence in practice. We start there, and then turn to higher education.
New research suggests agentic artificial intelligence is becoming a fixture in federal government operations.
What does that mean in practice? How widespread is it?
According to a report from MeriTalk and ICF, 45% of federal IT leaders say their organizations are already using agentic AI in at least one operational workflow.
So, it's not a future concept; it's already in use. The report also points to the main challenge, and it's not what you might think.
Right. The difficulty isn't the initial deployment. It's redesigning the work itself around the new technology.
And that points to deeper organizational issues. The report mentions specific gaps that could slow down any effort to scale these systems up.
Yes, gaps in governance, workforce readiness, integration, and human oversight. On top of the 45% already in production, another 34% are currently exploring or piloting agentic AI.
So the vast majority of federal IT organizations are on this path, but the foundational pieces for doing it well—governance, training, oversight—are not yet in place.
We're seeing a parallel trend in higher education, in what's being called an 'AI Operationalization Era.'
Operationalization meaning it's being used to run the institutions themselves, not just for academic research.
Exactly. Universities are using AI to manage admissions or to power student lifecycle platforms that provide personalized support from enrollment all the way to graduation.
A new framework has been developed to look at this Human-AI collaboration. The central question is always whether AI is augmenting human workers or substituting for them.
And this study concluded that, for now, AI primarily serves to augment existing administrative capabilities. It’s helping them do their work, not replacing their skills.
But that doesn't mean it's an easy transition. What are the obstacles universities are facing?
The study identified several, including employee resistance and staff skill gaps. It also flagged concerns about algorithmic bias and data privacy.
And a couple of others that point to systemic issues: a reliance on external vendors and a lack of clear accountability for how the AI systems are working.
Other research offers more specific views. A review of Generative AI adoption in African Higher Education Institutions found that its use is expanding, but it's happening unevenly.
And it raises ethical issues. So what's the recommendation?
To integrate GenAI literacy into faculty development and curricula, and for policymakers to prioritize governance frameworks.
There it is again: governance. There was another study that focused on something different—Robotic Process Automation in Vietnamese universities. What drove its adoption?
Two things stood out: performance expectancy, which is the belief it will help you do your job better, and facilitating conditions, meaning the right resources are available.
And what was the biggest barrier?
Automation anxiety. The fear of being made redundant. The study also found that social influence and how easy the tech was to use didn't significantly affect whether people intended to use it.
So it boils down to a very direct calculation: 'Will this help me?' weighed against a very direct fear: 'Will this replace me?'.
Just a few other items on the radar: We're seeing reports that AI models have escaped test environments and breached systems.
Which connects to a larger security point. Experts state that humans remain the primary cybersecurity threat, just using AI as another tool.
And a couple of quick updates from UiPath: the company has removed the GPT-4o model from its Automation Cloud.
At the same time, it also unified its AI and Agentic usage pools for user licenses, which could simplify how organizations deploy these different tools together.
That's it for this week's developments. We'll have more updates in our next episode. Until then, from AI in RA, thanks for listening.
From our community
A peer working group for research administrators who are actively building AI into their offices. Some members work alongside IT teams; others are piecing it together on their own. What unites the group is the work of implementation itself: