Unified Theory of Acceptance and Use of Technology

• UTAUT is a comprehensive framework that integrates eight antecedents into four core constructs—performance expectancy, effort expectancy, social influence, and facilitating conditions—to predict technology acceptance.
• The model has been empirically validated across various domains such as e-government, healthcare IT, and AI, using methodologies like CFA and PLS-SEM to explain up to 70% of variance in user behavior.
• Extensions like UTAUT2 and contextual adaptations enhance its applicability by incorporating hedonic motivation, price value, habit, and localized factors to drive real-world technology adoption.

The Unified Theory of Acceptance and Use of Technology (UTAUT) is a structural framework originally proposed to synthesize and extend prior models of individual technology acceptance. UTAUT systematically models the determinants of user acceptance and use behavior for information systems, drawing together eight theoretical antecedents into a streamlined set of predictors. Since its introduction, UTAUT has been empirically validated and extended across numerous domains—including e-government, enterprise AI adoption, healthcare IT, e-learning, and generative AI—using both classical and advanced measurement methodologies.

1. Theoretical Foundation and Core Constructs

UTAUT (Venkatesh et al., 2003) posits that four principal constructs explain the majority of variance in Behavioral Intention (BI) to use technology and Actual Use Behavior (UB):

• Performance Expectancy (PE): The user’s belief that using the technology will yield gains in job or task performance.
• Effort Expectancy (EE): The degree of ease associated with the use of the technology.
• Social Influence (SI): The user’s perception that important others believe they should use the new technology.
• Facilitating Conditions (FC): The belief that organizational and technical infrastructure is available to support system use.

The core structural relationships can be formalized as:

$$\begin{aligned} \text{BI} &= \beta_{1}\,\text{PE} + \beta_{2}\,\text{EE} + \beta_{3}\,\text{SI} + \zeta_{1} \ \text{UB} &= \beta_{4}\,\text{BI} + \beta_{5}\,\text{FC} + \zeta_{2} \end{aligned}$$

These relationships are moderated by gender, age, experience, and voluntariness of use; i.e., path coefficients vary by demographic strata and usage context (Alshehri et al., 2012, Alshehri et al., 2013).

2. Measurement Models and Empirical Validation

UTAUT constructs are operationalized through reflective Likert-scale items targeting defined user beliefs concerning performance, effort, social, and infrastructural dimensions. Confirmatory factor analysis (CFA) and Partial Least Squares–Structural Equation Modeling (PLS-SEM) are standard methodologies to validate:

Construct	Typical Cronbach’s α	Composite Reliability (CR)	AVE
Performance Expectancy (PE)	0.75 – 0.96	0.79 – 0.96	0.69 – 0.94
Effort Expectancy (EE)	0.76 – 0.96	0.77 – 0.96	0.53 – 0.94
Social Influence (SI)	0.77 – 0.93	0.77 – 0.94	0.79 – 0.89
Facilitating Conditions (FC)	0.73 – 0.95	0.83 – 0.95	0.84 – 0.93

Discriminant validity for all constructs is generally empirically confirmed using Fornell–Larcker and HTMT criteria (Alshehri et al., 2012, Alrawashdeh et al., 2012, Ajao et al., 14 Oct 2024). The model explains up to 70% of the variance in technology use behavior in large-sample SEM applications (Alshehri et al., 2012, Alshehri et al., 2013).

3. Structural Pathways, Moderation, and Key Findings

Empirical evidence across sectors robustly supports PE, EE, and FC as significant positive determinants of behavioral intention, with PE and FC often dominant (Alrawashdeh et al., 2012, Alshehri et al., 2012, Ajao et al., 14 Oct 2024, Badrawani, 13 Jun 2025). Social Influence is context-dependent: significant in education (Thongsri et al., 2019), non-significant in some e-government (Alshehri et al., 2012, Alshehri et al., 2013), and subject to reconceptualization in non-Western professional settings as "horizontal" peer pressure (Murtuza et al., 14 Nov 2025). Moderators such as Internet Experience can amplify path coefficients, e.g., strengthening EE→BI and FC→BI (Alshehri et al., 2013).

Empirical path estimates for typical adoption scenarios:

Context	PE → BI/USE	EE → BI/USE	SI → BI/USE	FC → BI/USE
E-Government (Saudi Arabia)	0.34	0.54	0.042 (n.s.)	0.38
E-Government with Website Quality	0.34	0.39	-0.03 (n.s.)	0.48
QRIS Mobile Payments (Indonesia)	0.096	0.096	0.185	0.173
EV Adoption (Nigeria; FC extended)	0.25	—	0.15	0.44

Where reported, FC becomes the single strongest predictor in resource-constrained or infrastructurally challenged contexts (Ajao et al., 14 Oct 2024). In AI adoption, affective factors (anxiety, attitude, self-efficacy) and organizational status show additional, albeit small, influences on both intention and usage intensity (Wolfe et al., 16 Oct 2025).

4. Model Extensions and Integration

a. UTAUT2

UTAUT2 (Venkatesh, Thong, Xu 2012) generalizes the original framework to consumer and post-adoption scenarios by adding:

• Hedonic Motivation (HM)
• Price Value (PV)
• Habit (HT)

Empirical studies show that in LLM adoption for software engineering, PE and habit are the only significant direct drivers of behavioral intention; HM and FC primarily influence actual use rather than intention (Lambiase et al., 8 Sep 2024). In generative AI adoption for journalists, "voluntary-compulsion" and the distinction between vertical and horizontal social influence are critical theoretical refinements (Murtuza et al., 14 Nov 2025).

b. Contextual/Cultural Localizations

Several studies extend UTAUT in light of local enablers:

• Facilitating Conditions: Extended as second‑order constructs (e.g., with infrastructure, policy, and affordability dimensions in Nigeria for EV adoption) (Ajao et al., 14 Oct 2024).
• New Domains: Domain-specific constructs such as "Website Quality" (Alshehri et al., 2012), “System Flexibility” and “System Enjoyment” (Alrawashdeh et al., 2012), and content realism or novelty in creative AI (Mvondo et al., 7 May 2024).
• Crisis and Policy Environments: Separation of policy signal and force-majeure (pandemic) risk in payment adoption (Badrawani, 13 Jun 2025).

c. Integration with Other Models

Healthcare IT research frequently integrates UTAUT with the Task–Technology Fit (TTF) model, introducing constructs for system "fit" with clinical tasks. The extended model posits that TTF mediates the effects of technology/task characteristics and is a strong determinant of behavioral intention, sometimes exceeding classic UTAUT predictors in thematic prominence (Aljarboa et al., 2020, Aljarboa et al., 2020).

5. Methodological Innovations

LLMs now enable rapid, robust annotation of unstructured user-generated content according to UTAUT constructs, producing scalable, survey-equivalent datasets with inter-annotator reliability comparable to human experts (Smolinski et al., 30 Jun 2024). Configurational methods such as fsQCA are used alongside variance-based SEM to reveal alternate sufficient conditions for adoption (Mvondo et al., 7 May 2024).

In leading-edge extensions, affective constructs—such as warm-glow, anxiety, self-efficacy, and attitude—are systematically introduced and validated to capture intrinsic and extrinsic motivational factors that classical cognitive-only models omit (Saravanos et al., 2022, Wolfe et al., 16 Oct 2025). These affective antecedents can become significant, surpassing classical predictors under specific motivational primes.

6. Theoretical and Practical Implications

The UTAUT framework provides a highly generalizable, modular basis for modeling technology acceptance, supporting:

• Comparative analyses across cultural, infrastructural, and regulatory contexts.
• Empirical quantification and decomposition of drivers in regulated, crisis, or resource-constrained environments.
• Integration with system-task fit, habit formation, affective reward, and localized determinants.

From a practical perspective, enhancing performance benefits, ensuring robust infrastructural and support conditions, leveraging peer-based social proof, and reducing affective barriers (e.g., anxiety) have all emerged as effective strategies to drive adoption across heterogeneous settings (Alshehri et al., 2012, Ajao et al., 14 Oct 2024, Wolfe et al., 16 Oct 2025). The adaptability of UTAUT to domain- and context-specific extensions underpins its durability as the dominant modeling approach in contemporary technology acceptance research.