Two tensions kept sharpening across this week’s signals. The first is evidence of learning: faculty are moving past whether students used AI toward a harder question, which is what actually shows that a student reasoned, revised, evaluated, and understood. The second is workforce breadth: AI fluency is no longer concentrated in computer science, and students across business, communications, psychology, biology, and economics are building AI experience while institutions are still deciding how much of that preparation belongs in the curriculum. The developments below suggest higher education is starting to respond at the level of course design, academic programs, faculty development, and infrastructure.
“A finished product tells you something was made. It does not tell you who did the thinking. That is the gap AI just made impossible to ignore.”
— Dr. Lynn Austin
EDUCAUSE Reframes the AI Assessment Question: From Student Products to Evidence of Capability
The Details
Tope Onitiri argues in EDUCAUSE Review that generative AI is exposing a limitation that predates ChatGPT: faculty often infer learning from finished products while much of the actual intellectual development stays invisible (Onitiri, 2026). A polished essay, presentation, or project can show what was produced without revealing how much reasoning, revision, or decision-making the student performed. AI-supported work can add evidence when students are asked to document what they questioned, accepted, rejected, reconsidered, and learned. Onitiri names critical thinking, metacognition, judgment, adaptability, ethical reasoning, and continued learning as capabilities that develop through repeated practice, and suggests faculty make that development visible by asking what evidence changed a student’s position, which AI recommendation they rejected and why, and what uncertainty remains.
Why it Matters
This changes assessment design without turning faculty into AI investigators. The emphasis moves from proving how a document was produced to gathering enough evidence to judge what the student understands. It also names a workload reality: institutions that expect instructors to verify learning have to help redesign assessments, rubrics, and course processes so verification is built into normal teaching rather than bolted on only when misconduct is suspected.
Handshake Finds AI Experience Spreading Well Beyond Computer Science Majors
The Details
Handshake’s 2026 Class of What’s Next report analyzed 12.4 million U.S. candidate profiles from the classes of 2024 through 2030, measured as of June 2026 (Handshake, 2026). Nearly two-thirds of candidates who document AI-related experience are not computer science majors, and even after including computing, engineering, and math, more than a quarter come from outside those fields. The most common nontechnical majors are business, economics, biology, communications, psychology, and marketing. Coursework remains a strong indicator of AI foundations, but students are also building skills through projects and certifications, and AI-titled internship postings draw nearly five times as many applications as postings that do not mention AI, from candidates with comparable qualifications.
Why it Matters
AI workforce preparation can no longer sit mainly inside computing departments. Business, communications, healthcare, social science, and education programs need to define what AI competence looks like in their own professions. The report also exposes a timing problem: students are learning emerging practices on their own when coursework does not address them, so departments need enough contact with employers and professional practice to decide which capabilities belong in courses and which tools are too transient to build into learning outcomes.
Policy & Governance
-
Research suggests AI can widen differences in how deeply students think
A mixed-methods study in built-environment education reports that AI use does not affect every learner in the same way: some students use AI to extend exploration while others rely on it in ways that reduce intellectual engagement, which complicates policies that treat all AI use as educationally equivalent (Crolla et al., 2026).
-
South Dakota State prepares AI-supported oral assessment pilots
South Dakota State University is expanding course-specific AI tutoring and plans to pilot AI-supported oral assessment in 2026-27. Faculty keep responsibility for grading while an AI agent can question students about course content, adding a way to assess depth of understanding beyond written work. Provost Dennis Hedge frames oral assessment as a way to gauge how well students grasp material and to make testing less centered on writing (South Dakota State University, 2026).
-
Stony Brook builds critical AI evaluation into the first-year experience
Stony Brook University Libraries announced on August 21 that information literacy and critical AI evaluation are being built into SBU 101 and ADV 101, so incoming students learn to evaluate scholarly information, sources, and Google AI Overviews rather than encountering AI literacy only when an individual instructor chooses to address it (Fena, 2026).
Programs, Research & Infrastructure
-
Stony Brook creates a department devoted to technology, AI, and society
Stony Brook University named Donghee Yvette Wohn inaugural chair of its new Department of Technology, AI and Society on August 17. Launched with a $5 million SUNY grant over three years, the department pairs AI research and education with work on societal impact, responsible AI, human-computer interaction, and communication (Stony Brook University, 2026).
-
Louisiana Tech pairs new AI research funding with dedicated computing
Louisiana Tech University announced a $599,982 National Science Foundation award to assistant professor Md Rubel Ahmed for work on translating AI models into specialized hardware, as its College of Engineering and Science prepares to open a privately funded 15-station NVIDIA AI laboratory in Nethken Hall this fall, connecting faculty research, student instruction, and infrastructure in one plan rather than treating each as a separate investment (Louisiana Tech University, 2026).
-
George Mason convenes universities and regional educators on AI readiness
George Mason University reported on August 19 on an ERA-NOVA convening that brought university faculty, school-system leaders, and community partners together on AI literacy, professional learning, policy, classroom practice, and workforce preparation. Participants prioritized modular professional development, train-the-trainer models, and stronger alignment between policy and instruction (George Mason University, 2026).
-
Faculty Focus frames verification as assessment, not investigation
B. Jean Mandernach outlines five approaches faculty can use when a submitted product does not provide enough evidence of learning, including short conversations, student videos, draft trails, and AI interaction logs. The useful distinction: verification asks students to demonstrate understanding and can be designed into assessment, without requiring an instructor to prove how a questionable submission was produced (Mandernach, 2026).
-
Florida State expands cross-campus AI preparation through a bootcamp series
Florida State University reported on August 21 on an AI Bootcamp series for faculty, staff, graduate students, and undergraduates across disciplines, drawing on computing, libraries, data science, information technology, chemistry, information science, national laboratories, industry, and other universities (Florida State University, 2026).
Do It Now Checklist
Betting On: Visible Learning
This week’s core lesson is that AI makes the process of learning more important, not less. Institutions need assessments and curricula that make reasoning, judgment, revision, and disciplinary understanding visible, while preparing students to use AI in the work they will actually encounter.
At Inspiration Moments, we share ideas meant to help you make deliberate choices for a more purposeful future. Betting on visible learning means making sure the evidence behind a grade shows what a student can think, explain, evaluate, and defend. Life happens for you, not to you, to live your purpose.
Respectfully,
Dr. Lynn Austin