📈 economics

Behavioral Pathways Behind Students’ Willingness to Pay for GenAI in Higher Education

This study utilizes the Technology Acceptance Model to demonstrate that undergraduate students' willingness to pay for Generative AI tools in higher education is driven by perceived ease of use and usefulness, which foster behavioral intentions that ultimately translate into financial commitment.

Original authors: Suraksha Baral, Sudarshan Adhikari, Emily Buck, Prerana Khatiwada

Published 2026-07-08

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Suraksha Baral, Sudarshan Adhikari, Emily Buck, Prerana Khatiwada

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Why Would Students Pay for AI?

Imagine Generative AI (like ChatGPT) as a new, magical kitchen gadget. It can chop vegetables, mix sauces, and even write recipes for you. Right now, most students are using the "free trial" version of this gadget. They love it, but they aren't paying for it.

This study asked a simple question: What makes a student decide to pull out their wallet and pay a monthly fee for this gadget?

The researchers didn't just ask, "Do you like it?" They looked at the mental steps a student takes before they are willing to spend money. They used a classic psychology framework called the Technology Acceptance Model (TAM), which is like a recipe for how people decide to use new technology.

The Three-Step "Payment Pathway"

The study found that students don't just wake up and decide to pay. They go through a specific chain reaction, like dominoes falling. Here is how it works:

1. The "Easy Button" (Perceived Ease of Use)

The Concept: Is the gadget hard to figure out? Does it require a manual, or can you just pick it up and start cooking?
The Finding: If the AI tool feels smooth, simple, and intuitive, students are more likely to think, "Hey, this is actually useful."
Analogy: Think of a new video game controller. If the buttons are sticky and confusing, you won't think the game is fun. But if the controller feels like an extension of your hand, you start to believe the game is worth playing.

2. The "Value Check" (Perceived Usefulness)

The Concept: Does this tool actually help me get my homework done faster or better?
The Finding: This is the most important step. Students only pay if they feel the tool is genuinely helping their grades or saving them time.
Analogy: Just because a car is easy to drive (Step 1) doesn't mean you'll buy it. You have to believe it will get you to work on time and save you gas money (Step 2). The study found that "ease of use" only leads to "willingness to pay" if it first convinces the student that the tool is actually useful.

3. The "Commitment" (Behavioral Intention)

The Concept: Do I plan to keep using this tool next week, next month, and next semester?
The Finding: Once students see the tool is easy and useful, they form a plan to keep using it. This plan is the bridge to paying.
Analogy: This is like deciding to join a gym. You try the equipment (Easy), you realize it helps you get fit (Useful), and then you decide, "I'm going to go here every Tuesday" (Intention).

The Final Result: Willingness to Pay

The Finding: Only after the student has decided they will keep using the tool do they consider paying for it.
The Reality Check: The study found a huge gap. While students love using these tools, only about 43% said they would pay for them. The average amount they were willing to pay was $5.71 a month.
The Problem: Real premium subscriptions for tools like ChatGPT Plus, Grammarly, and GitHub Copilot cost way more than $5.71 combined. It's like students saying, "We love this car, but we'll only pay $5 a month for gas, even though it costs $40."

What the Study Doesn't Say (and what it actually says)

It doesn't say: "Universities should force students to pay."
- It actually says: Students are currently dependent on free versions. If universities want students to pay, the tools must prove they are indispensable.
It doesn't say: "AI is perfect."
- It actually says: Students are willing to pay only when the tool reliably improves their work. If the tool is glitchy or hard to use, they won't pay.
It doesn't say: "This applies to everyone."
- It actually says: This study looked at 121 undergraduate students at one university in the US. It's a snapshot of a specific group, not a rule for the whole world.

The Takeaway

Think of this study as a map of a student's wallet. It shows that students aren't paying for AI because it's "cool" or "easy." They are paying because they have fallen into a routine where the tool is essential to their success.

The path to a student's wallet looks like this:

It's easy to use $\rightarrow$
It helps me learn $\rightarrow$
I plan to use it forever $\rightarrow$
Okay, I'll pay for it.

If any of those links in the chain are broken (for example, if the tool is easy but doesn't actually help with homework), the student will happily stick with the free version. The study suggests that for universities and companies to get students to pay, they need to focus on making the tools so useful that students feel they can't do their schoolwork without them.

Technical Summary: Behavioral Pathways Behind Students' Willingness to Pay for GenAI in Higher Education

Problem Statement
While Generative Artificial Intelligence (GenAI) is rapidly integrating into higher education, existing literature predominantly focuses on usage intentions or self-reported adoption. These metrics fail to capture the depth of student commitment, particularly regarding financial investment. Unlike institutionally licensed technologies (e.g., Learning Management Systems) or supplementary tools, GenAI occupies an ambiguous space characterized by generative adaptability, algorithmic opacity, and dual-tier ecosystems (free vs. premium). Consequently, there is a critical gap in understanding the mechanisms that translate student perceptions of GenAI into a willingness to pay (WTP). This study addresses the lack of systematic evidence regarding how students evaluate the economic value of GenAI tools and whether their behavioral intentions to use these tools translate into actual financial commitment.

Methodology
The study employs a quantitative, cross-sectional design utilizing the Technology Acceptance Model (TAM) extended to include economic outcomes.

Sample: Data was collected via an online survey from 121 undergraduate students at a large public university in the United States. The sample included a majority of female participants (72.7%) with a mean age of 20.78 years. Most participants (93.4%) had prior experience with GenAI, primarily ChatGPT.
Instrument: The survey measured four latent constructs based on TAM: Perceived Ease of Use (PEU), Perceived Usefulness (PU), Behavioral Intention (BI), and Willingness to Pay (WTP). PEU, PU, and BI were measured using multi-item 5-point Likert scales adapted from prior literature. WTP was measured as a single-item concrete monetary value (monthly USD), framed as the maximum amount students would pay if free access were removed.
Analysis: Data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). This approach was selected for its suitability in predictive modeling and its ability to handle complex mediation pathways with the given sample size. The analysis assessed indicator reliability, internal consistency, convergent validity, and discriminant validity (using Fornell-Larcker criteria and HTMT ratios) before testing the structural model via bootstrapping (5,000 resamples).

Key Results
The structural model analysis yielded significant findings supporting the proposed hypotheses:

Direct Effects:
- PEU → PU: Perceived ease of use positively and significantly predicted perceived usefulness ( $\beta = 0.61, p < .01$ ).
- PU → BI: Perceived usefulness strongly predicted behavioral intention to use GenAI ( $\beta = 0.73, p < .01$ ).
- BI → WTP: Behavioral intention positively predicted willingness to pay ( $\beta = 0.45, p < .01$ ).
Mediation Effects:
- The study confirmed that PEU does not directly influence WTP. Instead, its effect is fully mediated through PU and BI.
- H4: PU mediates the relationship between PEU and BI.
- H5: BI mediates the relationship between PU and WTP.
- H6: A serial mediation path exists where PEU influences WTP through the chain of PU and BI ( $\beta = 0.20, p < .01$ ).
Model Fit and Explanatory Power:
- The model demonstrated good fit (SRMR = 0.062).
- PEU explained 35.2% of the variance in PU.
- PU explained 53.4% of the variance in BI.
- BI explained 20.4% of the variance in WTP.
- All endogenous constructs showed positive predictive relevance ( $Q^2 > 0$ ).
Descriptive Findings:
- Despite high usage rates, only 43% of participants expressed a willingness to pay.
- The mean WTP was $5.71 per month, significantly lower than the cumulative cost of premium versions of common tools (e.g., ChatGPT Plus, Grammarly, Copilot), indicating a reliance on free tiers or institutional subsidies.

Key Contributions

Theoretical Extension: The study extends the Technology Acceptance Model (TAM) by incorporating Willingness to Pay as a downstream behavioral outcome in the context of GenAI. It validates the sequential logic of TAM (PEU $\to$ PU $\to$ BI $\to$ WTP) even when the outcome shifts from adoption to economic commitment.
Mechanism Clarification: It elucidates that ease of use alone is insufficient to drive payment; it must first enhance perceptions of usefulness, which then fosters behavioral intention, ultimately leading to financial commitment.
Empirical Evidence: It provides one of the first empirical examinations of the gap between GenAI usage and payment willingness in higher education, highlighting the specific economic dynamics of dual-tier AI ecosystems.

Significance and Claims
The paper claims that its findings offer critical guidance for universities, AI service providers, and developers regarding pricing and governance policies.

For Institutions: The results suggest that institutional subscriptions should be negotiated before students develop individual payment dependencies. Relying on students to pay individually post-adoption may fail due to low WTP and the "anchoring effect" of free access. Furthermore, institutions must provide clear guidelines on ethical use, as responsible use frameworks may influence the decision to pay for premium, reliable versions.
For Developers: Pricing models must align with student economic realities. The study indicates that students are willing to pay only when tools demonstrate clear, sustained academic value that justifies the cost beyond free alternatives. Improvements in accuracy, personalization, and workflow integration are necessary to convert usage intention into payment.
Equity Implications: The authors caution that without deliberate institutional action to address pricing misalignment, the transition to paid GenAI features risks widening educational inequalities, as free tiers may stagnate while premium capabilities offer superior learning support.

The study concludes that while TAM remains a robust framework for understanding GenAI adoption, the unique characteristics of GenAI (generativity, opacity, and dual-tier access) require specific policy interventions to ensure equitable access and sustainable adoption.

The Big Picture: Why Would Students Pay for AI?

The Three-Step "Payment Pathway"

What the Study Doesn't Say (and what it actually says)

The Takeaway

Technical Summary: Behavioral Pathways Behind Students' Willingness to Pay for GenAI in Higher Education

More like this