Case 2: The Adoption Decision — Innovare AI Analysis

As you read — hold this question

TAM predicts whether individuals will adopt AI. TOE predicts whether organisations are ready to adopt AI. If both frameworks say adoption should happen — but it doesn't — what did both models miss?

2 models

that explain AI adoption from different levels. TAM explains the individual. TOE explains the organisation. Neither explains the gap between them.

Technology adoption frameworks exist because adoption is not automatic. Useful technologies get rejected; inferior technologies achieve dominance. The Technology Acceptance Model (TAM) and the Technology-Organisation-Environment (TOE) framework are the two most widely applied frameworks in AI adoption research. TAM, developed in 1989, focuses on individual psychology: what makes a person willing to use a new tool? TOE, developed in 1990, focuses on organisational readiness: what conditions allow an organisation to successfully adopt? Together they describe the terrain. But the gap between "should adopt" and "actually adopted" is wider than either model alone can explain — and that gap is where most AI programmes fail.

Technology Acceptance Model

Quiz: TAM / TOE

TAM — what makes an individual adopt a new technology

Fred Davis and Richard Bagozzi (1989) developed TAM to understand why individuals accept or reject information technology. The model rests on two core factors that predict attitude, which in turn predicts intention, which predicts actual use.

Technology Acceptance Model (TAM)

Davis & Bagozzi, 1989

Perceived Usefulness — the degree to which a person believes using the system will enhance their job performance. "Will this make me better at what I do?" If the answer is unclear, adoption will be slow regardless of ease of use.

PEOU

Perceived Ease of Use — the degree to which a person believes using the system will be effortless. "Will this create more work than it saves?" A powerful but complex tool can fail on PEOU alone.

Flow

PU + PEOU → Attitude → Behavioural Intention → Actual Use. The chain is sequential: you can intervene at any stage, but you cannot skip one.

Applied to AI

What this means in practice

The AI equivalent of PU is value clarity: employees must be able to see a specific benefit to their own work, not the organisation in general. "AI will transform our industry" does not create adoption. "This tool will draft your weekly report in 8 minutes instead of 90" does.

The AI equivalent of PEOU is friction: how many extra steps does using AI add to existing workflow? Each additional step reduces adoption. Integration into existing tools (rather than a separate platform people must log into) dramatically lowers the barrier.

Technology-Organisation-Environment Framework

TOE — what makes an organisation ready to adopt

Tornatzky, Fleischer and Chakrabarti (1990) developed the TOE framework to explain technology adoption at the organisational level. Where TAM looks inward (individual psychology), TOE looks at context: what internal and external conditions determine whether adoption is feasible?

Technological Context — the AI itself

Functionality: how well the AI performs its intended tasks · Complexity: ease or difficulty of implementing and using it · Compatibility: how well it integrates with existing systems · Relative Advantage: perceived benefits over what is currently used. An AI that scores poorly on any of these creates friction before a single employee has touched it.

Organisational Context — internal conditions

Size and Structure: larger organisations have more resources but slower decision cycles · Culture: organisations that support innovation and change have an inherent adoption advantage · Top Management Support: commitment from senior leadership is not optional; it is the single factor most correlated with successful AI implementation · Resource Availability: financial, human, and technical resources to actually sustain the initiative after launch.

Environmental Context — external pressures

Market Conditions: competitive pressure can force adoption regardless of internal readiness · Regulatory Environment: regulations can either enable or block adoption outright · Technological Infrastructure: cloud computing and connectivity are prerequisites; without them, AI cannot run · Social and Cultural Norms: public perception of AI affects how customers and staff respond to its deployment.

TOE predicts adoption readiness. An organisation that scores well across all three contexts — clear technology fit, strong internal culture and leadership, supportive external environment — is well positioned to adopt. But readiness is not adoption. The gap between them is where TAM becomes relevant again.

What both models miss

The gap between "should adopt" and "actually adopted"

TAM and TOE are descriptive frameworks: they explain the conditions under which adoption should happen. They do not explain what to do when it doesn't. The empirical gap between readiness and actual adoption is explained by factors neither model fully captures.

Gap factor	TAM coverage	TOE coverage	What's actually needed
Psychological safety	Not addressed	Partial (culture)	Employees must feel safe disclosing AI use and mistakes. Fear of punishment suppresses adoption even when PU and PEOU are high.
Trust in AI outputs	Implied by PU	Not addressed	If employees don't trust the AI's outputs, they will double-check everything — negating the efficiency gain and destroying perceived usefulness.
Incentive alignment	Not addressed	Not addressed	If sharing AI use leads to workload increases without recognition, rational employees hide their use. Reward systems must change.
Individual experimentation	Implicit	Not addressed	AI has a jagged frontier of capability. Only the person doing the task can discover which AI outputs are useful for their specific work. Top-down mandates cannot substitute.

The Missing Piece

Both TAM and TOE assume adoption is the goal and describe conditions for it. Neither asks why a fully capable, willing employee in a supportive organisation might still choose not to adopt. The answer — explored in Case 6 — is that the incentives often point away from adoption, and no framework score can override a rational self-interest calculation.

Take this away

TAM explains the individual; TOE explains the organisation. Neither explains the decision an individual makes about whether it is safe, rewarding, and rational to adopt within their specific organisational context. That gap is where most AI programmes fail — not on technology, not on organisational readiness, but on the human calculation of risk versus reward.

← Case 1: The VUCA World

Case 3: The Leadership Gap →

Quick recall — without looking back

Test yourself on this case

Question 1 of 3

What are the two core factors in TAM, and how do they combine to predict whether an individual will adopt AI?

Perceived Usefulness (PU): the degree to which someone believes AI will enhance their job performance — "will this make me better at what I do?" Perceived Ease of Use (PEOU): the degree to which someone believes using AI will be effortless — "will this create more work than it saves?" Both factors shape Attitude toward the technology, which drives Behavioural Intention, which leads to Actual Use. Both factors must be addressed: high PU but high complexity (low PEOU) can still result in rejection; and an easy tool with no perceived benefit will also be rejected.

Question 2 of 3

Name the three contexts in the TOE framework. For each, give two specific factors that influence AI adoption.

Technological context: Functionality (how well the AI performs its tasks) and Compatibility (how well it integrates with existing systems). Also: Complexity and Relative Advantage. Organisational context: Top Management Support (leadership commitment, cited as the single most critical factor) and Organisational Culture (innovation-supportive cultures adopt more readily). Also: Size/Structure and Resource Availability. Environmental context: Market Conditions (competitive pressure can force adoption regardless of readiness) and Regulatory Environment (regulations can enable or block adoption). Also: Technological Infrastructure and Social/Cultural Norms.

Question 3 of 3

TAM and TOE can both score an organisation as "ready to adopt." Why might adoption still not happen?

Both models describe conditions for adoption but do not address the individual's risk calculation within the organisation. Key gaps include: psychological safety (employees may fear disclosure of AI use even when culture is nominally supportive); trust in AI outputs (if employees don't trust the model's outputs, PEOU suffers regardless of the tool's actual ease of use); incentive misalignment (if sharing AI use leads to more work without recognition, hiding it is rational); and individual experimentation (AI has a jagged capability frontier — only the individual doing the task can discover what works for their specific use case, and top-down mandates cannot substitute for this).

Module 4 Video

Module 4 · Video Walkthrough

Sources

Davis & Bagozzi: Davis, F.D. & Bagozzi, R.P. (1989). User acceptance of computer technology: a comparison of two theoretical models. Management Science, 35(8), 982–1003.
Tornatzky et al.: Tornatzky, L.G., Fleischer, M. & Chakrabarti, A.K. (1990). The Processes of Technological Innovation. Lexington Books.
Module transcript: BUSN9049 Module 4 Part 2 — AI acceptance and technology adoption models. Flinders University, 2026.
Module slides: BUSN9049 Module 4 — Challenges in AI Implementation. Flinders University, 2026.