Artificial Intelligence Adoption in SMEs: Survey Based on TOE–DOI Framework, Primary Methodology and Challenges
Abstract
:1. Introduction
2. AI’s Benefits for Business and Peculiarities for SMEs
3. Critical Analysis of AI Adoption in SMEs: Internal and External Factors Through the TOE–DOI Framework
3.1. TOE Plus DOI Framework Relevant to SMEs
3.2. Technological Determinant
3.3. Organizational Determinant
3.4. Environmental Determinant
3.5. Empirical Findings on SMEs and AI Adoption
4. Empirical Insights from Sectorial and Regional Studies
4.1. AI Adoption in Manufacturing and Retails SMEs
4.1.1. AI-Driven Smart Manufacturing: Technological and Organizational Lessons from Customization Models [109]
4.1.2. AI and IoT Integration in Retail and Manufacturing SMEs: Operational Efficiencies and Strategic Challenges [110]
4.2. AI Synergies and Revenue Growth in European SMEs
4.2.1. AI Adoption in European SMEs and the Role of Internal Capabilities [27]
4.2.2. A Comprehensive Systematic Literature Review [112]
4.2.3. AI and Robotics as Innovation Drivers in European SMEs [113]
4.3. Nordic Countries Study for AI Adoption in Manufacturing and Retail SMEs
4.4. AI Adoption and Digital Competencies in Spanish SMEs
4.5. Insights from Italian SMEs: Organizational Readiness and Strategic Implementation
4.6. AI-Driven Business Model Innovation for Sustainability: Evidence from U.S. SMEs
4.7. AI-Driven Bank Digitalization and SME Financing in China
4.8. Digital Readiness as a Foundation for AI Adoption in Emerging Market SMEs in India
4.9. Saudi Arabia: Sustainable Business Performance Through AI Adoption
4.10. AI and Financial Inclusion in African SMEs: A Fintech-Driven Perspective
4.11. AI Adoption and Sustainable Performance in Developing Economies: Evidence from Pakistani SMEs
4.12. AI Marketing and SME Performance in Emerging Economies: Evidence from Ghana SMEs
4.13. Summary Table to Complement the Cases Analysis
5. Challenges and Actionable Pathways for AI Adoption in SMEs
5.1. Closing the Knowledge Gap and Technical Expertise
- A.
- Develop Internal Capabilities through Training. Invest in continuous role-specific training for employees in areas such as data analysis, AI fundamentals, and algorithmic thinking. Training should be modular and accessible, allowing gradual learning and practical application in daily operations.
- B.
- Leverage Low-Code and Pre-Trained AI Tools. Adopt user-friendly, cloud-based platforms offering pre-configured AI models. These reduce the technical entry barrier and enable SMEs to gain value from AI without requiring advanced programming or data science skills.
- C.
- Build Strategic External Partnerships. Collaborate with universities, research centers, AI consulting firms, and technology providers. These partnerships provide access to expert guidance, reduce learning curves, and help SMEs co-develop solutions tailored to their business context.
- D.
- Engage in Open Innovation Networks. Participate in innovation ecosystems and industry alliances to exchange knowledge, gain exposure to best practices, and remain informed about AI trends and tools suitable for SME environments.
5.2. Designing Scalable AI Solutions for Sustainable Growth
- A.
- Adopt Modular Implementation ApproachesBegin with small-scale, manageable AI projects that are aligned with specific business needs. This reduces upfront risk while providing opportunities to validate performance and refine processes prior to full-scale deployment.
- B.
- Utilize Agile MethodologiesEmploy agile project management techniques that support iterative development and regular feedback. This increases responsiveness to business changes and helps align AI solutions with evolving operational priorities.
- C.
- Leverage Scalable Cloud PlatformsUse cloud-based infrastructures that enable on-demand computing power and storage. This allows SMEs to scale AI applications gradually, reducing the need for heavy capital investments in IT hardware.
- D.
- Ensure Future Integration CompatibilitySelect AI tools and architectures that are interoperable and adaptable to future technologies or systems. This forward-thinking approach avoids vendor lock-in and allows easier expansion and system upgrades.
5.3. Overcoming Financial Barriers to Enable AI Adoption
- A.
- Explore Flexible Financing MechanismsSMEs can access external financing options such as public grants, low-interest innovation loans, and tax incentives. Shared-resource models (e.g., AI-as-a-Service or pay-as-you-go platforms) also lower entry costs and reduce upfront capital expenditure.
- B.
- Implement Pilot Tests to Reduce RiskPilot projects allow SMEs to test AI solutions in limited controlled environments before making full-scale investments. This phased approach provides early evidence of ROI and builds organizational confidence while minimizing financial exposure.
- C.
- Adopt Cloud-Based and Scalable SolutionsUsing cloud infrastructure allows SMEs to avoid high infrastructure costs. Pay-per-use models allow businesses to expand computing resources in line with their growth and adoption pace, reducing the need for costly on-premise systems.
- D.
- Foster Public–Private CollaborationPartnering with governmental agencies, technology vendors, and innovation hubs can open access to subsidies, co-investment schemes, and technical expertise. These alliances reduce the burden of individual investment and create supportive adoption ecosystems.
5.4. Ensuring Data Availability, Quality, and Governance for AI Success
- A.
- Establish Robust Data Collection SystemsIdentify key internal and external data sources and implement continuous real-time data collection mechanisms. This improves the quantity and relevance of the data for model training and decision-making.
- B.
- Improve Data Quality through Cleaning and ValidationInvest in systematic data cleaning processes, removing duplicates, correcting errors, and filling gaps, to ensure the integrity and usability of datasets. High-quality data are essential to generate accurate and reliable AI output.
- C.
- Facilitate Data Integration across SystemsUse APIs and cloud platforms to unify data from disparate systems, including legacy platforms. Integrated data ecosystems prevent information silos and support scalable AI solutions that align with firm operations.
- D.
- Leverage Data Enrichment and Analytics ToolsApply data mining, augmentation, and visualization tools to expand the scope and depth of available data. This enhances the insight generation capacity of AI models and supports more informed business decisions.
- E.
- Implement Strong Data Governance PracticesDefine and enforce policies on data access, privacy, storage, and compliance. Align governance with national and international regulations (e.g., GDPR) to ensure ethical, transparent, and secure data usage.
5.5. Building Future-Ready Infrastructure and Integration Pathways
- A.
- Adopt Cloud-Based AI Platforms Leverage scalable cloud infrastructure to reduce initial investment and ensure on-demand access to computing resources.
- B.
- Collaborate with Public–Private AI Centers Partner with national or regional AI innovation hubs to access shared infrastructure, expert support and high-performance computing tools.
- C.
- Use APIs for System InteroperabilityDeploy APIs to bridge legacy systems and new AI tools, allowing SMEs to avoid full system overhauls while improving compatibility.
- D.
- Invest in Modular Digital InfrastructureDesign infrastructure upgrades in modular phases to support future AI applications without disrupting current operations.
5.6. Cultivating an Innovation-Driven Culture to Support AI Adoption
- A.
- Promote an Innovation Mindset InternallyEncourage creativity, openness to new ideas, and willingness to experiment. Support innovation champions within the organization and reward proactive behavior that contributes to digital initiatives.
- B.
- Implement AI through Gradual, Pilot-Driven ApproachesUse small-scale pilot projects to introduce AI incrementally, allowing employees to gain confidence, understand its value and see its alignment with long-term goals without overwhelming disruption.
- C.
- Create Internal Spaces for Learning and ExperimentationEstablish safe environments for trial-and-error, where staff can test ideas and learn new tools without fear of failure. This fosters continuous learning and reduces resistance to change.
- D.
- Encourage Open Innovation and External CollaborationEngage employees, clients, partners, and external experts in co-creation. Opening innovation processes to diverse stakeholders increases idea generation and fosters a sense of shared ownership over change.
- E.
- Lead Change Management with Transparency and InclusionClearly communicate the purpose, benefits, and implications of AI adoption. Address concerns directly, involve staff in solution design, and promote social responsibility along with economic efficiency.
5.7. Enhancing Human–AI Collaboration to Improve Organizational Productivity
- Centaur teams represent human–machine pairings that jointly perform tasks to outperform either work alone.
- Cyborg workers use embedded or connected AI systems to amplify their own capabilities in real-time, increasing performance and adaptability.
- A.
- Design AI systems that complement—not replace—human rolesFocus on tools that support decision-making, improve accuracy, or automate repetitive tasks while leaving critical thinking and judgment to employees.
- B.
- Invest in employee upskilling for AI collaborationProvide training on how to work effectively with AI tools and interpret AI-generated outputs to empower employees in augmented workflows.
- C.
- Promote a co-intelligence mindset across the organizationEncourage leadership and teams to frame AI not as a substitute, but as a partner in productivity and innovation.
- D.
- Experiment with augmentation models such as Centaur and Cyborg approachesPilot use cases where human–AI synergy can be observed, measured, and refined, especially in customer service, product development, or strategic planning.
5.8. Strengthening Public–Private Collaboration to Support AI Adoption
- A.
- Develop Inclusive Innovation EcosystemsFoster structured partnerships among governments, technology providers, universities, and research centers. These ecosystems should offer shared resources, such as cloud access, computing infrastructure, training programs, and co-investment platforms.
- B.
- Establish Collaborative Spaces for SMEs and Tech StakeholdersCreate physical and virtual environments, such as AI hubs, open innovation labs and online platforms, where SMEs can co-develop AI solutions, share experiences, and get advice from technical experts and researchers.
- C.
- Implement Tailored Financial IncentivesDesign public funding mechanisms such as grants, tax credits, and subsidized loans specifically targeted at SMEs seeking AI adoption. Encourage private sector co-financing through equity partnerships or innovation vouchers.
- D.
- Promote Knowledge Exchange and Joint Capacity BuildingLaunch joint research initiatives, AI-focused training programs, and cross-sector mentorships that connect SMEs with academic and corporate expertise. These initiatives can reduce skill gaps and accelerate learning.
- E.
- Introduce Adaptive and Innovation-Friendly RegulationSimplify regulatory procedures and promote standards that support experimentation and reduce compliance burdens. This fosters a more agile environment for testing and adopting AI technologies.
5.9. Strategically Leveraging Generative AI to Democratize Capabilities and Drive Innovation: Toward Open-Weight LLM Adoption
- A.
- Adopt Accessible, Low-Code Gen-AI PlatformsBegin with intuitive, commercially available Gen-AI tools that support content creation, communication, ideation, and task automation. Low-code platforms such as Jasper, Notion AI, Copy.ai, ChatGPT, and Canva Magic Write offer SMEs immediate entry points to experiment with AI without needing specialized technical skills. These tools require minimal onboarding, typically follow a free-mium pricing model, and can be embedded into existing workflows.
- B.
- Enhance Workforce Capabilities through AugmentationUse Gen-AI to augment, not replace, human capabilities. Empower employees across departments (for example, marketing, customer service, R&D) to harness Gen-AI to write reports, generate insights, answer internal queries, or accelerate ideation. This augmentation model reinforces a co-intelligence approach where AI assists, but human judgment remains central.To accelerate adoption, SMEs can introduce ’AI pilots’ within departments to pilot tools and coach others. Platforms such as ChatGPT for Business, Microsoft Copilot, or Google Workspace AI integrate directly with standard office software, making the augmentation process seamless. In parallel, free and paid learning modules on platforms such as LinkedIn Learning, Coursera, or Google’s AI Academy help build internal capability.
- C.
- Align Gen-AI with Strategic and Innovation GoalsIntegrate Gen-AI into broader strategic planning and innovation processes. Rather than treating AI as a side experiment, SMEs should embed Gen-AI into their OKRs (Objectives and Key Results), innovation roadmaps, and performance KPIs. Use cases may include trend analysis, scenario simulation, customer journey visualization, and stakeholder communication support.For example, SMEs can use Gen-AI to draft strategic memos, synthesize competitor reports, or explore “what-if” analyses related to product launches or market entry. Alignment with strategic goals ensures that AI investments are not just technically sound, but also business-relevant. Decision support functions such as executive briefings, sales forecasting, or risk scenario modeling are particularly impactful.
- D.
- Leverage Small Open-Weight LLMs for Flexible and Scalable DeploymentFor SMEs with intermediate or advanced digital maturity, small open-weight LLMs (e.g., Mistral-7B, LLaMA 3, DeepSeek-R1, Phi-2) present an attractive alternative to API-based tools. These models offer higher transparency, control over fine-tuning, and deployment cost-efficiency—particularly relevant for privacy-sensitive industries or firms with unique use cases.Modular deployments can begin in containerized environments using tools such as Docker, Kubernetes, and libraries such as Hugging Face Transformers, LangChain, or vLLM. These support inference at the edge or on affordable GPU cloud services (e.g., Paperspace, RunPod, or Replicate).SMEs can gradually scale from local pilots to fully integrated Gen-AI pipelines, ensuring that each step aligns with internal governance, data sovereignty, and customization needs. Use cases such as internal knowledge bases, structured summarization, or semantic search are ideal starting points for in-house LLM hosting.
- E.
- Accessible Catalog of Practical Applications: A Strategic Ecosystem ImperativeSMEs would benefit significantly from access to a curated and regularly updated catalog of Generative AI applications, particularly those based on LLMs, that showcase real-world use cases and documented productivity improvements. Such a catalog should include sector-specific and cross-functional case studies, benchmarks, deployment guides, and integration templates—providing SMEs with actionable insights on how GenAI can enhance efficiency, reduce costs, and unlock new forms of innovation.At present, no widely recognized repository exists that systematically compiles and contextualizes successful GenAI applications tailored for SMEs. This absence creates a significant barrier, as many SMEs lack the internal capacity to explore the breadth of AI solutions or to translate abstract capabilities into operational use cases. A well-designed catalog would act as a practical bridge—demystifying AI by showing how similar-sized firms have applied it for marketing automation, customer support, internal documentation, compliance analysis, or knowledge management.Establishing and maintaining such a catalog should be viewed as an ecosystem-level priority for policy actors, industry alliances, and AI providers. Without this, SMEs will continue to face limited access to replicable adoption pathways—hampering both innovation scalability and cross-sector learning.Illustrative Case Example: A mid-sized logistics SME in Spain integrated Mistral-7B via Hugging Face and FastAPI to automate client email summarization and route planning queries. This reduced human workload by approximately 40% and achieved a 12% increase in customer satisfaction within six weeks. Hosting was implemented through a GPU instance on Paperspace, reducing infrastructure costs by 35% compared to equivalent proprietary API services.As needs evolve, transitioning to open-weight LLMs [122], such as DeepSeek-R1, LLaMA, Mistral, or Falcon, offers SMEs greater autonomy and transparency. These models allow for self-hosting, fine-tuning, or deployment on private infrastructure, making them suitable for SMEs operating under strict compliance, cost, or customization requirements. This flexibility enables more sustainable, ethically guided, and strategically aligned AI adoption.
- F.
- Scale Strategically from Pilots to Core FunctionsBegin with low-risk, high-frequency use cases—such as drafting marketing copy, summarizing meeting notes, automating internal documentation, or creating customer FAQs—where Gen-AI can deliver immediate efficiency gains with minimal risk. These entry points allow SMEs to experiment with generative technologies, build organizational confidence, and fine-tune deployment practices in a contained environment.As internal trust and technical familiarity increase, expand AI applications to more complex and value-adding business functions, including customer service automation (e.g., chatbot integration), sales intelligence, supply chain prediction, or even product design ideation. Each expansion phase should be guided by clearly defined success metrics (e.g., response time reduction, conversion uplift, error rate minimization) and paired with stakeholder feedback loops.Strategic scaling requires a governance layer that ensures alignment with business priorities, ethical safeguards, and compliance constraints. This may involve forming an internal AI task force, defining data usage policies, and setting thresholds for human-in-the-loop oversight.Example Pathway:
- –
- Phase 1—Exploration: Deploy Gen-AI for internal document drafting and idea generation.
- –
- Phase 2—Operational Integration: Extend Gen-AI to handle customer emails, generate marketing briefs, or automate routine analysis.
- –
- Phase 3—Strategic Embedding: Integrate Gen-AI into CRM systems, product development workflows, or knowledge management platforms.
A phased, feedback-driven rollout mitigates risk, nurtures AI maturity across the organization, and enables SMEs to evolve from exploratory use to strategic integration—maximizing return on investment and long-term competitiveness. - G.
- Addressing Practical Integration and Mitigation of LLM RisksWhile open-weight LLMs offer SMEs greater autonomy, transparency, and cost control compared to closed platforms, their deployment introduces non-trivial risks that must be strategically managed. Common technical challenges include prompt injection attacks, model hallucination (i.e., generation of false or misleading outputs), and the propagation of encoded biases. These risks are further compounded by SMEs’ typically limited in-house AI expertise and constrained cybersecurity budgets.In practice, SMEs must weigh the flexibility of self-hosting against the operational overhead it entails—such as provisioning GPU-based infrastructure, managing version control, ensuring uptime, and enforcing data security protocols. Without sufficient safeguards, improperly configured LLMs can produce unreliable results or expose sensitive business data to unintended uses.Recommended Mitigations and Integration Strategies:
- –
- Lightweight Guardrails: Implement tools for content moderation, such as prompt sanitization, response length limits, and input validation layers. Libraries such as Guardrails.ai, LangChain safety chains, or Rebuff offer plug-and-play solutions to mitigate harmful output and misuse.
- –
- Use of Containerized Frameworks: Leverage containerized deployments (e.g., Docker with Hugging Face Transformers + FastAPI) to simplify integration and enforce isolation. This reduces the complexity of managing dependencies and accelerates reproducibility.
- –
- Monitoring and Feedback Loops: Incorporate human-in-the-loop (HITL) oversight for critical tasks and establish monitoring tools (e.g., logging, real-time review dashboards) to capture hallucination incidents or drift in model behavior over time.
- –
- Shared Maintenance Communities: Rely on open communities and maintained libraries (e.g., OpenLLM, LLMGuard, or Hugging Face Model Hub) that offer security patches, pretrained safety layers, and updated weights. These ecosystems provide a valuable lifeline for SMEs without dedicated AI teams.
- –
- Prioritized Use Cases: Limit LLM deployment to high-ROI, low-risk applications—such as internal knowledge search, summarization, and document generation—before expanding to customer-facing or compliance-critical contexts.
- –
- Ethical and Compliance Guidelines: Develop internal guidelines addressing transparency, data provenance, and consent in GenAI applications. These should reflect emerging standards (e.g., EU AI Act (https://artificialintelligenceact.eu/, accessed on 12 May 2025), ISO/IEC 42001 (https://www.iso.org/standard/81230.html, accessed on 12 May 2025)) and foster a culture of responsible use.
Ultimately, adopting open-weight LLMs is not merely a technical decision but a governance challenge. For SMEs, combining lightweight safety protocols with community-supported tooling and a clear risk assessment process is essential to avoid harm, preserve trust, and build sustainable AI capabilities. - H.
- Establish Ethical Governance and Usage GuidelinesIntroduce lightweight but effective ethical governance structures tailored to the size and capabilities of SMEs. These frameworks should include principles of transparency, fairness, accountability, and security in the use of Generative AI (Gen-AI), particularly large language models (LLMs). Clear documentation of how Gen-AI tools are used internally (e.g., for content creation or decision support) and externally (e.g., customer interactions) helps build trust with employees, clients, and regulatory bodies.Transparency guidelines should ensure that Gen-AI-generated content is identifiable, especially in customer-facing contexts. This includes labeling AI-generated responses or content to avoid misleading stakeholders. Bias mitigation practices must be implemented to reduce the risk of replicating harmful stereotypes or skewed outputs in automated content. This can be achieved using prompt calibration, fairness-aware fine-tuning, and output filtering tools.SMEs should also establish internal policies outlining appropriate use cases, limitations of Gen-AI tools, employee responsibilities, and data protection protocols. These policies should align with emerging legal frameworks such as the EU AI Act or ISO/IEC 42001, ensuring compliance with data privacy, accountability, and algorithmic impact assessment norms.To promote accountability, SMEs can assign a responsible AI lead or committee—even if informally composed—who periodically reviews Gen-AI use, tracks incidents (e.g., hallucinations, misuse), and updates governance practices as technologies evolve. Training staff on these principles reinforces an ethical culture and reduces misuse risk.Ultimately, ethical governance is not an overhead but a strategic enabler—providing guardrails that reduce legal exposure, foster user confidence, and ensure Gen-AI serves as a sustainable and trusted innovation driver.
5.10. Embedding Responsible AI Practices and Human-Centered Governance
- A.
- Integrate Ethical Considerations into AI PlanningBegin AI projects with structured reflection on potential impacts, such as bias, discrimination, or data misuse. Embed human rights principles into AI decision-making from the outset, even in pilot projects.
- B.
- Establish Lightweight Governance StructuresWhile SMEs may lack formal ethics boards, simple mechanisms, such as designating an AI responsibility lead or incorporating stakeholder reviews, can support ethical oversight and build internal accountability.
- C.
- Promote Transparency and ExplainabilityEnsure that AI-generated outputs (e.g., decisions, recommendations) are explainable to nontechnical users and customers. Use interfaces that allow humans to override, question, or validate AI suggestions.
- D.
- Strengthen Data Protection and Consent ProtocolsAdhere to national and international data privacy laws. Where possible, ensure that AI systems use consent-based anonymized data and that data handling policies are clearly communicated.
- E.
- Engage Stakeholders in Responsible InnovationInvolve employees, users, and partners in discussions about AI risks, values, and priorities. This participatory approach fosters shared responsibility and aligns AI solutions with organizational values.
5.11. Mapping AI Adoption Challenges and Solutions Within the TOE Framework
6. Structured Pimary Methodology for Effective AI Adoption in SMEs
- 1.
- Conduct a Comprehensive Readiness AssessmentEvaluate the organizational technological infrastructure, digital maturity, workforce competencies, leadership alignment, and financial capacity to establish a realistic starting point for AI adoption.
- 2.
- Define Strategic Objectives and Use CasesSet measurable AI-related goals related to business priorities, such as improving efficiency, improving customer experience, or enabling data-driven decisions.
- 3.
- Select Appropriate AI Solutions and Deployment ModelsChoose between off-the-shelf, customized, or cloud-based AI tools based on internal needs, available resources, and scalability requirements.
- 4.
- Implement in Phases with Pilot Projects Begin with small-scale pilots to validate the feasibility and value of AI, make necessary adjustments and build internal trust and capacity before greater deployment.
- 5.
- Integrate Training and Change ManagementEnsure ongoing staff training and transparent communication to address resistance, clarify roles and reinforce the value of AI as an enhancement tool.
- 6.
- Measure, Adjust, and Scale StrategicallyEstablish key performance indicators (KPI) to track progress. Use insights from early phases to iterate and scale AI solutions in broader operations.
7. Discussion: Lessons Learned, Limitations, and Future Research Directions
“AI adoption is no longer optional for SMEs; it has become a strategic lever for transformation and a core requirement to remain competitive in a digital and increasingly regulated marketplace.”
7.1. Lessons Learned
- AI presents transformative opportunities for SMEs.AI enables automation, smarter decision-making, customer personalization, and new value creation pathways. When aligned with the lean structures and constraints of SMEs, these tools can significantly enhance competitiveness and innovation (TOE: Technological).
- Adoption lags due to multi-dimensional barriers.Despite its potential, AI uptake in SMEs remains slow. Challenges include technological complexity, financial constraints, lack of digital culture, and limited external support. The TOE–DOI framework helps identify and address these systematically at the structural and perceptual levels.
- Adoption strategies must reflect SME-specific realities.SMEs operate with unique restrictions, limited resources, agile decision-making, and informal structures. Effective adoption of AI requires tailored and scalable approaches sensitive to these characteristics (TOE: Organizational).
- AI must be embedded in strategy—not treated as a side project.AI should be part of the firm’s core strategic planning, not an isolated technology experiment. This requires long-term vision, structured roadmaps, and alignment of leadership. Our six-phase implementation model offers practical guidance for SMEs to embed AI into their growth strategies (TOE: Organizational + Technological).
- Organizational culture is a key enabler.A culture that supports experimentation, learning, and inclusive change management promotes smoother adoption of AI. Overcoming fear and resistance requires leadership to create safe spaces for adaptation (TOE: Organizational).
- Talent development and upskilling are essential.Technical and managerial knowledge of AI must be built internally. Upskilling existing staff and reducing dependence on external consultants enhances sustainability and internal capability (TOE: Organizational).
- Accessible, modular, and open-weight solutions can democratize adoption.Cloud-based platforms, plug-and-play tools, and open-weight LLMs—such as LLaMA, DeepSeek-R1, Mistral, and Falcon—offer SMEs scalable, transparent, and cost-effective alternatives to proprietary systems. These models support self-hosted, hybrid, or cloud-based deployment, enabling greater control over data, compliance, and customization. By lowering technical and financial barriers, open-weight solutions empower SMEs to adopt AI in ways that align with their strategic goals and operational realities (TOE: Technological + Organizational).
- Data quality and governance underpin successful AI.Reliable, accessible, and clean data is the foundation for effective AI. SMEs must invest in data infrastructure, governance protocols, and integration tools such as APIs to ensure actionable insights (TOE: Technological).
- Ethical and responsible AI must be integrated from the start.As SMEs adopt powerful AI tools, governance mechanisms around fairness, transparency, and data protection are essential. Lightweight and actionable frameworks can help SMEs ensure trust and compliance (TOE: Organizational + Environmental).
- Public-private ecosystems should support SME AI adoption.Collaboration with universities, governments, tech providers and AI computation centers can provide SMEs with access to expertise, infrastructure and financial support, especially in early adoption phases (TOE: Environmental).
7.2. Limitations and Future Research
- First, the proposed challenge mapping matrix (Table 4) is constructed from secondary literature and has not yet been validated through primary data collection or empirical fieldwork.
- Second, the findings are context sensitive and may not generalize uniformly across all industries, regions, or firm sizes, particularly given the rapidly evolving nature of generative AI technologies and regulatory landscapes.
- Third, the framework assumes a relatively linear adoption trajectory, whereas real-world implementation processes are often iterative, adaptive, and path-dependent.
- The proposed model would benefit from empirical validation across sectors and geographies. Although the framework is grounded in robust literature, its operationalization should be tested through case studies, surveys, and longitudinal studies. As shown by [123], AI assimilation affects firm performance via dynamic capabilities such as absorptive capacity and customer agility. Future research could explore how these mediators vary between different levels of digital maturity and organizational size.
- Studies such as [6,13] highlight structural disparities in AI adoption between SMEs and larger enterprises. These include differences in project management integration, support levels, and digital readiness. Comparative analyzes and longitudinal research could uncover which internal or contextual factors allow certain SMEs to overcome common adoption bottlenecks such as infrastructure gaps, talent shortages, and strategic misalignment.
- Strategic alignment remains an underexplored area. Refs. [115,124] propose structured adoption roadmaps tailored to SMEs, yet these models require broader testing across industries and regions. Future work could examine how such phased approaches perform under varying degrees of resource constraint, leadership involvement, or policy support.
- Generative AI (Gen-AI) tools represent a transformative opportunity, particularly for smaller firms. However, as noted in [3,119], there is limited empirical data on how SMEs integrate Gen-AI into workflows or innovation cycles. Future studies should investigate adoption pathways, productivity outcomes, and organizational learning processes associated with Gen-AI tools such as ChatGPT, Copilot, or Jasper. Furthermore, exploring how Gen-AI supports sustainability goals, as discussed by [3], could guide AI strategies oriented to climate in SMEs.
- Implementing ethical and responsible AI remains another urgent area. Ref. [67] calls for lightweight governance models that can be embedded in SMEs with limited regulatory capacity. More research should explore how such models can be institutionalized, including stakeholder engagement practices, explainability protocols, and ethical impact audits, especially in environments with emerging or fragmented regulatory oversight.
- Technology-wise, infrastructure readiness and integration continue to be critical barriers, as detailed in [97,112]. Future research should address modular, cloud-based, and API-driven architectures that support interoperability and scalability, especially for resource-constrained SMEs in emerging economies.
- In addition, studies at the policy and ecosystem level could build on and [100], investigating the role of public–private partnerships, national digital strategies, and innovation funding to support the AI adoption journeys of SMEs. Exploring how government interventions affect regional disparities and inclusive innovation will help craft context-specific recommendations for digital policy frameworks.
- Finally, as proposed by [125], the evolving taxonomy of AI applications in innovation management provides a rich avenue to study how SMEs use AI in innovation of product, process and business model. Mapping these applications to industry verticals and innovation stages can offer valuable granularity to existing adoption models.
8. Conclusions
Author Contributions
Funding
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- World Bank. Small and Medium Enterprises (SMEs) Finance: Improving SMEs’ Access to Finance and Finding Innovative Solutions to Unlock Sources of Capital; World Bank: Washington, DC, USA, 2024. [Google Scholar]
- Dvorsky, J.; Kubalek, J.; Barinova, D.; Androniceanu, A.; Khan, K.A. Selected Factors Influencing the Social and Environmental Aspects of Sustainability of Smes. Transform. Bus. Econ. 2024, 23, 62. [Google Scholar]
- Shaik, A.S.; Alshibani, S.M.; Jain, G.; Gupta, B.; Mehrotra, A. Artificial intelligence (AI)-driven strategic business model innovations in small-and medium-sized enterprises. Insights on technological and strategic enablers for carbon neutral businesses. Bus. Strategy Environ. 2024, 33, 2731–2751. [Google Scholar] [CrossRef]
- Raj, K.; Fredrick, D.P.; Kurahattidesai, C.; Hegde, C.S. Artificial Intelligence Driven Customer Relationship Management: Harnessing the power of technology to improve business efficiency. Int. J. Commun. Netw. Inf. Secur. 2024, 16, 58–65. [Google Scholar]
- Pingali, S.R.; Singha, S.; Arunachalam, S.; Pedada, K. Digital readiness of small and medium enterprises in emerging markets: The construct, propositions, measurement, and implications. J. Bus. Res. 2023, 164, 113973. [Google Scholar] [CrossRef]
- Schwaeke, J.; Peters, A.; Kanbach, D.K.; Kraus, S.; Jones, P. The new normal: The status quo of AI adoption in SMEs. J. Small Bus. Manag. 2024, 63, 1–35. [Google Scholar] [CrossRef]
- Costa, J.; Castro, R. SMEs must go online—E-commerce as an escape hatch for resilience and survivability. J. Theor. Appl. Electron. Commer. Res. 2021, 16, 3043–3062. [Google Scholar] [CrossRef]
- Yatna, C.N.; Banjuradja, S.Y.; Sutrisno, S.; Ritawaty, N. Digital Technology Adaptation Challenges to Enhance Growth and Security of Corporate Online Businesses. Adv. Bus. Ind. Mark. Res. 2025, 3, 44–56. [Google Scholar] [CrossRef]
- Baabdullah, A.M. The precursors of AI adoption in business: Towards an efficient decision-making and functional performance. Int. J. Inf. Manag. 2024, 75, 102745. [Google Scholar] [CrossRef]
- Füller, J.; Hutter, K.; Wahl, J.; Bilgram, V.; Tekic, Z. How AI revolutionizes innovation management–Perceptions and implementation preferences of AI-based innovators. Technol. Forecast. Soc. Change 2022, 178, 121598. [Google Scholar] [CrossRef]
- Hansen, E.B.; Bøgh, S. Artificial intelligence and internet of things in small and medium-sized enterprises: A survey. J. Manuf. Syst. 2021, 58, 362–372. [Google Scholar] [CrossRef]
- Chatterjee, S.; Chaudhuri, R.; Grandhi, B.; Galati, A. Evolution of strategy for global value creation in MNEs: Role of knowledge management, technology adoption, and financial investment. J. Int. Manag. 2023, 29, 101057. [Google Scholar] [CrossRef]
- Tominc, P.; Oreški, D.; Čančer, V.; Rožman, M. Statistically Significant Differences in AI Support Levels for Project Management between SMEs and Large Enterprises. AI 2024, 5, 136–157. [Google Scholar] [CrossRef]
- Ghobakhloo, M.; Iranmanesh, M.; Vilkas, M.; Grybauskas, A.; Amran, A. Drivers and barriers of Industry 4.0 technology adoption among manufacturing SMEs: A systematic review and transformation roadmap. J. Manuf. Technol. Manag. 2022, 33, 1029–1058. [Google Scholar] [CrossRef]
- Hwang, W.S.; Kim, H.S. Does the adoption of emerging technologies improve technical efficiency? Evidence from Korean manufacturing SMEs. Small Bus. Econ. 2022, 59, 627–643. [Google Scholar] [CrossRef]
- Enholm, I.M.; Papagiannidis, E.; Mikalef, P.; Krogstie, J. Artificial intelligence and business value: A literature review. Inf. Syst. Front. 2022, 24, 1709–1734. [Google Scholar] [CrossRef]
- Lamarre, E. Digital and AI Transformations, Start with the Problem, Not the Technology; McKinsey & Company: New York, NY, USA, 2023. [Google Scholar]
- Tornatzky, L.; Fleischer, M. The Process of Technology Innovation; D.C. Heath & Company: Lexington, MA, USA, 1990. [Google Scholar]
- Rogers, E. Diffusion of Innovations; The Free Press: New York, NY, USA, 2003. [Google Scholar]
- Kumar Bhardwaj, A.; Garg, A.; Gajpal, Y. Determinants of blockchain technology adoption in supply chains by small and medium enterprises (SMEs) in India. Math. Probl. Eng. 2021, 2021, 5537395. [Google Scholar] [CrossRef]
- Das, S.D.; Bala, P.K. What drives MLOps adoption? An analysis using the TOE framework. J. Decis. Syst. 2024, 33, 376–412. [Google Scholar] [CrossRef]
- Simina, M.M.; Dutescu, A. TOE framework elements used on Artificial Intelligence implementation in the accounting and audit sector. Int. J. Res. Bus. Soc. Sci. 2024, 13, 335–349. [Google Scholar]
- Chen, H.; Li, L.; Chen, Y. Explore success factors that impact artificial intelligence adoption on telecom industry in China. J. Manag. Anal. 2021, 8, 36–68. [Google Scholar] [CrossRef]
- Rawash, H.N. E-commerce adopting TOE model by SMEs in Jordan. Multicult. Educ. 2021, 7, 118–122. [Google Scholar]
- Chatterjee, S.; Rana, N.P.; Dwivedi, Y.K.; Baabdullah, A.M. Understanding AI adoption in manufacturing and production firms using an integrated TAM-TOE model. Technol. Forecast. Soc. Change 2021, 170, 120880. [Google Scholar] [CrossRef]
- El-Haddadeh, R. Digital innovation dynamics influence on organisational adoption: The case of cloud computing services. Inf. Syst. Front. 2020, 22, 985–999. [Google Scholar] [CrossRef]
- Arroyabe, M.F.; Arranz, C.F.; De Arroyabe, I.F.; de Arroyabe, J.C.F. Analyzing AI adoption in European SMEs: A study of digital capabilities, innovation, and external environment. Technol. Soc. 2024, 79, 102733. [Google Scholar] [CrossRef]
- Polisetty, A.; Chakraborty, D.; Kar, A.K.; Pahari, S. What determines AI adoption in companies? Mixed-method evidence. J. Comput. Inf. Syst. 2024, 64, 370–387. [Google Scholar] [CrossRef]
- Baabdullah, A.M.; Alalwan, A.A.; Slade, E.L.; Raman, R.; Khatatneh, K.F. SMEs and artificial intelligence (AI): Antecedents and consequences of AI-based B2B practices. Ind. Mark. Manag. 2021, 98, 255–270. [Google Scholar] [CrossRef]
- Ocloo, C.E.; Xuhua, H.; Akaba, S.; Shi, J.; Worwui-Brown, D.K. The determinant factors of business to business (B2B) E-commerce adoption in small-and medium-sized manufacturing enterprises. J. Glob. Inf. Technol. Manag. 2020, 23, 191–216. [Google Scholar] [CrossRef]
- Hamad, W.B. The role of ICT in knowledge management processes: A Review. Int. J. Eng. Sci. Comput. 2018, 8, 16373–16380. [Google Scholar]
- Sharma, S. Benefits or concerns of AI: A multistakeholder responsibility. Futures 2024, 157, 103328. [Google Scholar] [CrossRef]
- Huang, M.H.; Rust, R.T. A framework for collaborative artificial intelligence in marketing. J. Retail. 2022, 98, 209–223. [Google Scholar] [CrossRef]
- Badet, J. AI, automation and new jobs. Open J. Bus. Manag. 2021, 9, 2452–2463. [Google Scholar] [CrossRef]
- Umutoni, A. The Influence of Artificial Intelligence on Customer Service Automation in E-Commerce in Rwanda. Int. J. Technol. Syst. 2025, 10, 57–68. [Google Scholar] [CrossRef]
- Alami, H.; Rivard, L.; Lehoux, P.; Hoffman, S.J.; Cadeddu, S.B.M.; Savoldelli, M.; Samri, M.A.; Ag Ahmed, M.A.; Fleet, R.; Fortin, J.P. Artificial intelligence in health care: Laying the foundation for responsible, sustainable, and inclusive innovation in low-and middle-income countries. Glob. Health 2020, 16, 1–6. [Google Scholar] [CrossRef] [PubMed]
- Briatore, F.; Mosca, M.T.; Mosca, R.N.; Braggio, M. A Bibliometric Analysis on Artificial Intelligence in the Production Process of Small and Medium Enterprises. AI 2025, 6, 54. [Google Scholar] [CrossRef]
- Leocádio, D.; Guedes, L.; Oliveira, J.; Reis, J.; Melão, N. Customer service with AI-powered human-robot collaboration (HRC): A literature review. Procedia Comput. Sci. 2024, 232, 1222–1232. [Google Scholar] [CrossRef]
- Shaalan, A.; Tourky, M.; Ibrahim, K. AI Caramba!: The Negative Effects of AI Agents in Customer Relationship Management. In Leveraging AI for Effective Digital Relationship Marketing; IGI Global: Hershey, PA, USA, 2025; pp. 309–352. [Google Scholar]
- Matosas-López, L. The influence of brand credibility and brand loyalty on customer satisfaction and continued use intention in new voice assistance services based on AI. J. Mark. Anal. 2025, 13, 180–201. [Google Scholar] [CrossRef]
- Bhuiyan, M.S. The role of AI-Enhanced personalization in customer experiences. J. Comput. Sci. Technol. Stud. 2024, 6, 162–169. [Google Scholar] [CrossRef]
- Konda, R. AI-driven customer support: Transforming user experience and operational efficiency. Int. J. Sci. Technol. 2025, 16, 117–130. [Google Scholar]
- Gowri, D.P. Impact of AI in Personalized Digital Marketing: Boosting Customer Engagement through Tailored Content. J. Commun. Manag. 2024, 3, 216–221. [Google Scholar] [CrossRef]
- Gupta, Y.; Khan, F.M. Role of artificial intelligence in customer engagement: A systematic review and future research directions. J. Model. Manag. 2024, 19, 1535–1565. [Google Scholar] [CrossRef]
- Biallas, M.; O’Neill, F. Artificial intelligence innovation in financial services. Int. Financ. Corp. 2020, 85, 1–8. [Google Scholar]
- Ben Ayed, R.; Hanana, M. Artificial intelligence to improve the food and agriculture sector. J. Food Qual. 2021, 2021, 5584754. [Google Scholar] [CrossRef]
- Debbadi, R.K.; Boateng, O. Optimizing end-to-end business processes by integrating machine learning models with UiPath for predictive analytics and decision automation. Int. J. Sci. Res. Arch. 2025, 14, 778–796. [Google Scholar] [CrossRef]
- Chaudhary, P.S.; Khurana, M.R.; Ayalasomayajula, M. Real-world applications of data analytics, big data, and machine learning. In Data Analytics and Machine Learning: Navigating the Big Data Landscape; Springer: Berlin/Heidelberg, Germany, 2024; pp. 237–263. [Google Scholar]
- Vittori, D.; Natalicchio, A.; Panniello, U.; Petruzzelli, A.M.; Albino, V.; Cupertino, F. Failure is an option: How failure can lead to disruptive innovations. Technovation 2024, 129, 102897. [Google Scholar] [CrossRef]
- Stettler, T.R.; Moosauer, E.J.; Schweiger, S.A.; Baldauf, A.; Audretsch, D. Absorptive capacity in a more (or less) absorptive environment: A meta-analysis of contextual effects on firm innovation. J. Prod. Innov. Manag. 2025, 42, 18–47. [Google Scholar] [CrossRef]
- Denicolai, S.; Zucchella, A.; Magnani, G. Internationalization, digitalization, and sustainability: Are SMEs ready? A survey on synergies and substituting effects among growth paths. Technol. Forecast. Soc. Change 2021, 166, 120650. [Google Scholar] [CrossRef]
- Nagy, M.; Lăzăroiu, G.; Valaskova, K. Machine intelligence and autonomous robotic technologies in the corporate context of SMEs: Deep learning and virtual simulation algorithms, cyber-physical production networks, and Industry 4.0-based manufacturing systems. Appl. Sci. 2023, 13, 1681. [Google Scholar] [CrossRef]
- Etienne Fabian, N.; Dong, J.Q.; Broekhuizen, T.; Verhoef, P.C. Business value of SME digitalisation: When does it pay off more? Eur. J. Inf. Syst. 2024, 33, 383–402. [Google Scholar] [CrossRef]
- Giuggioli, G.; Pellegrini, M.M. Artificial intelligence as an enabler for entrepreneurs: A systematic literature review and an agenda for future research. Int. J. Entrep. Behav. Res. 2023, 29, 816–837. [Google Scholar] [CrossRef]
- Dörr, L.; Fliege, K.; Lehmann, C.; Kanbach, D.K.; Kraus, S. A taxonomy on influencing factors towards digital transformation in SMEs. J. Small Bus. Strategy 2023, 33, 53–69. [Google Scholar] [CrossRef]
- Horani, O.M.; Al-Adwan, A.S.; Yaseen, H.; Hmoud, H.; Al-Rahmi, W.M.; Alkhalifah, A. The critical determinants impacting artificial intelligence adoption at the organizational level. Inf. Dev. 2023, 02666669231166889. [Google Scholar] [CrossRef]
- Gangwar, H.; Date, H.; Raoot, A. Review on IT adoption: Insights from recent technologies. J. Enterp. Inf. Manag. 2014, 27, 488–502. [Google Scholar] [CrossRef]
- Sun, Y.; Tan, C.W.; Lim, K.H.; Liang, T.P.; Yeh, Y.H. Strategic contexts, strategic orientations and organisational technology adoption: A configurational approach. Inf. Syst. J. 2024, 34, 1355–1401. [Google Scholar] [CrossRef]
- Maroufkhani, P.; Wan Ismail, W.K.; Ghobakhloo, M. Big data analytics adoption model for small and medium enterprises. J. Sci. Technol. Policy Manag. 2020, 11, 483–513. [Google Scholar] [CrossRef]
- Loo, M.K.; Ramachandran, S.; Raja Yusof, R.N. Systematic Review of Factors and Barriers Influencing E-Commerce Adoption among SMEs over the Last Decade: A TOE Framework Perspective. J. Knowl. Econ. 2024, 1–25. [Google Scholar] [CrossRef]
- Faiz, F. Factors Influencing Digital Technologies Adoption among Indonesian SMEs: A Conceptual Framework. In Proceedings of the International Conference on Entrepreneurship, Leadership and Business Innovation (ICELBI 2022); Atlantis Press: Dordrecht, The Netherlands, 2024; pp. 227–241. [Google Scholar]
- Kim, J.S.; Seo, D. Foresight and strategic decision-making framework from artificial intelligence technology development to utilization activities in small-and-medium-sized enterprises. Foresight 2023, 25, 769–787. [Google Scholar] [CrossRef]
- Prasad Agrawal, K. Towards adoption of generative AI in organizational settings. J. Comput. Inf. Syst. 2024, 64, 636–651. [Google Scholar] [CrossRef]
- Toufaily, E.; Zalan, T.; Dhaou, S.B. A framework of blockchain technology adoption: An investigation of challenges and expected value. Inf. Manag. 2021, 58, 103444. [Google Scholar] [CrossRef]
- Effendi, M.I.; Sugandini, D.; Istanto, Y. Social media adoption in SMEs impacted by COVID-19: The TOE model. J. Asian Financ. Econ. Bus. 2020, 7, 915–925. [Google Scholar] [CrossRef]
- Abed, S.S. Social commerce adoption using TOE framework: An empirical investigation of Saudi Arabian SMEs. Int. J. Inf. Manag. 2020, 53, 102118. [Google Scholar] [CrossRef]
- Baldassarre, M.T.; Caivano, D.; Fernandez Nieto, B.; Gigante, D.; Ragone, A. Fostering Human Rights in Responsible AI: A Systematic Review for Best Practices in Industry. IEEE Trans. Artif. Intell. 2025, 6, 416–431. [Google Scholar] [CrossRef]
- Tiwari, A.K.; Marak, Z.R.; Paul, J.; Deshpande, A.P. Determinants of electronic invoicing technology adoption: Toward managing business information system transformation. J. Innov. Knowl. 2023, 8, 100366. [Google Scholar] [CrossRef]
- Merhi, M.I.; Harfouche, A. Enablers of artificial intelligence adoption and implementation in production systems. Int. J. Prod. Res. 2024, 62, 5457–5471. [Google Scholar] [CrossRef]
- Ebert, C.; Louridas, P. Generative AI for software practitioners. IEEE Softw. 2023, 40, 30–38. [Google Scholar] [CrossRef]
- Russo, D. Navigating the complexity of generative ai adoption in software engineering. ACM Trans. Softw. Eng. Methodol. 2024, 33, 1–50. [Google Scholar] [CrossRef]
- Chatterjee, S.; Chaudhuri, R. Adoption of artificial intelligence integrated customer relationship management in organizations for sustainability. In Business Under Crisis, Volume III: Avenues for Innovation, Entrepreneurship and Sustainability; Springer: Berlin/Heidelberg, Germany, 2022; pp. 137–156. [Google Scholar]
- Polas, M.R.H.; Jahanshahi, A.A.; Kabir, A.I.; Sohel-Uz-Zaman, A.S.M.; Osman, A.R.; Karim, R. Artificial intelligence, blockchain technology, and risk-taking behavior in the 4.0 IR metaverse era: Evidence from Bangladesh-based SMEs. J. Open Innov. Technol. Mark. Complex. 2022, 8, 168. [Google Scholar] [CrossRef]
- Choudrie, J.; Manandhar, N.; Castro, C.; Obuekwe, C. Hey Siri, Google! Can you help me? A qualitative case study of smartphones AI functions in SMEs. Technol. Forecast. Soc. Change 2023, 189, 122375. [Google Scholar] [CrossRef]
- von Garrel, J.; Jahn, C. Design framework for the implementation of AI-based (service) business models for small and medium-sized manufacturing enterprises. J. Knowl. Econ. 2023, 14, 3551–3569. [Google Scholar] [CrossRef]
- Kinney, M.; Anastasiadou, M.; Naranjo-Zolotov, M.; Santos, V. Expectation management in AI: A framework for understanding stakeholder trust and acceptance of artificial intelligence systems. Heliyon 2024, 10, e28562. [Google Scholar] [CrossRef]
- Grünbichler, R. Implementation barriers of artificial intelligence in companies. In Proceedings of the FEB Zagreb International Odyssey Conference on Economics and Business, Poreč, Croatia, 10–13 May 2023; Volume 5, pp. 193–203. [Google Scholar]
- Zavodna, L.S.; Überwimmer, M.; Frankus, E. Barriers to the implementation of artificial intelligence in small and medium-sized enterprises: Pilot study. J. Econ. Manag. 2024, 46, 331–352. [Google Scholar] [CrossRef]
- Yang, S.; Yang, S.; Tu, C.Y. Explore the drivers of e-commerce adoption by SMEs. Int. J. Bus. Manag. 2024, 4, 45. [Google Scholar] [CrossRef]
- Iansiti, M.; Lakhani, K.R. Competing in the Age of AI: Strategy and Leadership when Algorithms and Networks Run the World; Harvard Business Press: Cambridge, MA, USA, 2020. [Google Scholar]
- Lemos, S.I.; Ferreira, F.A.; Zopounidis, C.; Galariotis, E.; Ferreira, N.C. Artificial intelligence and change management in small and medium-sized enterprises: An analysis of dynamics within adaptation initiatives. Ann. Oper. Res. 2022, 1–27. [Google Scholar] [CrossRef] [PubMed]
- Chatterjee, S.; Chaudhuri, R.; Vrontis, D.; Basile, G. Digital transformation and entrepreneurship process in SMEs of India: A moderating role of adoption of AI-CRM capability and strategic planning. J. Strategy Manag. 2022, 15, 416–433. [Google Scholar] [CrossRef]
- Pellegrini, M.M.; Ciampi, F.; Marzi, G.; Orlando, B. The relationship between knowledge management and leadership: Mapping the field and providing future research avenues. J. Knowl. Manag. 2020, 24, 1445–1492. [Google Scholar] [CrossRef]
- Mladenova, I. SMEs in a Digital Era: The Role of Management. Adm. Sci. 2024, 14, 296. [Google Scholar] [CrossRef]
- Huseyn, M.; Ruiz-Gándara, Á.; González-Abril, L.; Romero, I. Adoption of artificial intelligence in small and medium-sized enterprises in Spain: The role of competences and skills. Amfiteatru Econ. 2024, 26, 848–866. [Google Scholar] [CrossRef]
- Weber, P. Unrealistic optimism regarding artificial intelligence opportunities in human resource management. Int. J. Knowl. Manag. (IJKM) 2023, 19, 1–19. [Google Scholar] [CrossRef]
- Liu, B. Integration of novel uncertainty model construction of green supply chain management for Small and Medium-Sized Enterprises using artificial intelligence. Optik 2023, 273, 170411. [Google Scholar] [CrossRef]
- Trawnih, A.; Yaseen, H.; Al-Adwan, A.S.; Alsoud, R.; Jaber, O.A. Factors influencing social media adoption among smes during COVID-19 crisis. J. Manag. Inf. Decis. Sci. 2021, 24, 1–18. [Google Scholar]
- Chaudhuri, R.; Chatterjee, S.; Vrontis, D.; Chaudhuri, S. Innovation in SMEs, AI dynamism, and sustainability: The current situation and way forward. Sustainability 2022, 14, 12760. [Google Scholar] [CrossRef]
- Boubaker, S.; Le, T.D.; Ngo, T.; Manita, R. Predicting the performance of MSMEs: A hybrid DEA-machine learning approach. Ann. Oper. Res. 2023, 1–23. [Google Scholar] [CrossRef]
- Dey, P.K.; Chowdhury, S.; Abadie, A.; Vann Yaroson, E.; Sarkar, S. Artificial intelligence-driven supply chain resilience in Vietnamese manufacturing small-and medium-sized enterprises. Int. J. Prod. Res. 2024, 62, 5417–5456. [Google Scholar] [CrossRef]
- Acemoglu, D. The simple macroeconomics of AI. Econ. Policy 2025, 40, 13–58. [Google Scholar] [CrossRef]
- Ganguly, K.K. Understanding the challenges of the adoption of blockchain technology in the logistics sector: The TOE framework. Technol. Anal. Strateg. Manag. 2024, 36, 457–471. [Google Scholar] [CrossRef]
- Sun, W.; Dedahanov, A.T.; Shin, H.Y.; Li, W.P. Using extended complexity theory to test SMEs’ adoption of Blockchain-based loan system. PLoS ONE 2021, 16, e0245964. [Google Scholar] [CrossRef] [PubMed]
- Omar, M.A.; Sulaiman, R.B. Social Commerce Adoption among MSME in Kuwait: The Role of Perceived Value and Organizational Innovation. Appl. Math. Inf. Sci. 2024, 18, 1105–1116. [Google Scholar]
- Areiqat, A.Y.; Alheet, A.F.; Qawasmeh, R.A.; Zamil, A.M. Artificial intelligence and its drastic impact on e-commerce progress. Acad. Strateg. Manag. J. 2021, 20, 1–11. [Google Scholar]
- Badghish, S.; Soomro, Y.A. Artificial intelligence adoption by SMEs to achieve sustainable business performance: Application of technology–organization–environment framework. Sustainability 2024, 16, 1864. [Google Scholar] [CrossRef]
- Fonseka, K.; Jaharadak, A.A.; Raman, M. Impact of E-commerce adoption on business performance of SMEs in Sri Lanka; Moderating role of artificial intelligence. Int. J. Soc. Econ. 2022, 49, 1518–1531. [Google Scholar] [CrossRef]
- Wei, R.; Pardo, C. Artificial intelligence and SMEs: How can B2B SMEs leverage AI platforms to integrate AI technologies? Ind. Mark. Manag. 2022, 107, 466–483. [Google Scholar] [CrossRef]
- Abrokwah-Larbi, K.; Awuku-Larbi, Y. The impact of artificial intelligence in marketing on the performance of business organizations: Evidence from SMEs in an emerging economy. J. Entrep. Emerg. Econ. 2024, 16, 1090–1117. [Google Scholar] [CrossRef]
- Hasani, T.; Rezania, D.; Levallet, N.; O’Reilly, N.; Mohammadi, M. Privacy enhancing technology adoption and its impact on SMEs’ performance. Int. J. Eng. Bus. Manag. 2023, 15, 18479790231172874. [Google Scholar] [CrossRef]
- Beebeejaun, A.; Gunputh, R.P. A study of the influence of artificial intelligence and its challenges: The impact on employees of the legal sector of Mauritius. Glob. Bus. Rev. 2023, 09721509231193803. [Google Scholar] [CrossRef]
- Bukartaite, R.; Hooper, D. Automation, artificial intelligence and future skills needs: An Irish perspective. Eur. J. Train. Dev. 2023, 47, 163–185. [Google Scholar] [CrossRef]
- Jones, M.; Idrovo-Carlier, S.; Rodriguez, A.J. Automation in Colombia: Assessing skills needed for the future of work. High. Educ. Ski. Work.-Based Learn. 2022, 12, 225–240. [Google Scholar] [CrossRef]
- Kinkel, S.; Baumgartner, M.; Cherubini, E. Prerequisites for the adoption of AI technologies in manufacturing—Evidence from a worldwide sample of manufacturing companies. Technovation 2022, 110, 102375. [Google Scholar] [CrossRef]
- Akpan, I.J.; Udoh, E.A.P.; Adebisi, B. Small business awareness and adoption of state-of-the-art technologies in emerging and developing markets, and lessons from the COVID-19 pandemic. J. Small Bus. Entrep. 2022, 34, 123–140. [Google Scholar] [CrossRef]
- Kalogiannidis, S.; Kalfas, D.; Loizou, E.; Papaevangelou, O.; Chatzitheodoridis, F. Smart sustainable marketing and emerging technologies: Evidence from the Greek business market. Sustainability 2023, 16, 312. [Google Scholar] [CrossRef]
- Msomi, T.; Olarewaju, O. Evaluation of access to finance, market and viability of small and medium-sized enterprises in South Africa. Probl. Perspect. Manag. 2021, 19, 281. [Google Scholar] [CrossRef]
- Wan, J.; Li, X.; Dai, H.N.; Kusiak, A.; Martinez-Garcia, M.; Li, D. Artificial-intelligence-driven customized manufacturing factory: Key technologies, applications, and challenges. Proc. IEEE 2020, 109, 377–398. [Google Scholar] [CrossRef]
- Haider, U.; Faisal, A. IoT and AI Synergy: Enhancing SMEs’ Operational Strategies in Retail and Manufacturing Sectors; ResearchGate GmbH: Berlin, Germany, 2024. [Google Scholar]
- Ardito, L.; Filieri, R.; Raguseo, E.; Vitari, C. Artificial intelligence adoption and revenue growth in European SMEs: Synergies with IoT and big data analytics. Internet Res. 2024. [Google Scholar] [CrossRef]
- Oldemeyer, L.; Jede, A.; Teuteberg, F. Investigation of artificial intelligence in SMEs: A systematic review of the state of the art and the main implementation challenges. Manag. Rev. Q. 2024, 75, 1185–1227. [Google Scholar] [CrossRef]
- Segarra-Blasco, A.; Tomàs-Porres, J.; Teruel, M. AI, robots and innovation in European SMEs. Small Bus. Econ. 2025, 1–27. [Google Scholar] [CrossRef]
- Peretz-Andersson, E.; Tabares, S.; Mikalef, P.; Parida, V. Artificial intelligence implementation in manufacturing SMEs: A resource orchestration approach. Int. J. Inf. Manag. 2024, 77, 102781. [Google Scholar] [CrossRef]
- Proietti, S.; Magnani, R. Assessing AI Adoption and Digitalization in SMEs: A Framework for Implementation. arXiv 2025, arXiv:2501.08184. [Google Scholar]
- Zhang, Y.; Yu, Y.; Chen, Z. AI, SME Financing, and Bank Digitalization. China Econ. Transit. Dangdai Zhongguo Jingji Zhuanxing Yanjiu 2022, 5, 210–241. [Google Scholar]
- Omokhoa, H.E.; Odionu, C.S.; Azubuike, C.; Sule, A.K. Digital transformation in financial services: Integrating AI, fintech, and innovative solutions for SME growth and financial inclusion. Gulf J. Adv. Bus. Res. 2024, 2, 423–434. [Google Scholar] [CrossRef]
- Soomro, R.B.; Al-Rahmi, W.M.; Dahri, N.A.; Almuqren, L.; Al-Mogren, A.S.; Aldaijy, A. A SEM–ANN analysis to examine impact of artificial intelligence technologies on sustainable performance of SMEs. Sci. Rep. 2025, 15, 5438. [Google Scholar] [CrossRef]
- Kshetri, N.; Rojas-Torres, D.; Hanafi, M.M.; Al-kfairy, M.; O’Keefe, G.; Feeney, N. Harnessing Generative Artificial Intelligence: A Game-Changer for Small and Medium Enterprises. IT Prof. 2025, 26, 84–89. [Google Scholar] [CrossRef]
- Mollick, E. Co-Intelligence: Living and Working with AI; Portfolio/Penguin: New York, NY, USA, 2024. [Google Scholar]
- Haase, J.; Klessascheck, F.; Mendling, J.; Pokutta, S. Sustainability via LLM Right-sizing. arXiv 2025, arXiv:2504.13217. [Google Scholar]
- Kukreja, S.; Kumar, T.; Purohit, A.; Dasgupta, A.; Guha, D. A literature survey on open source large language models. In Proceedings of the 2024 7th International Conference on Computers in Management and Business, Singapore, 12–14 January 2024; pp. 133–143. [Google Scholar]
- Wahab, M.D.A.; Radmehr, M. The impact of AI assimilation on firm performance in small and medium-sized enterprises: A moderated multi-mediation model. Heliyon 2024, 10, e29580. [Google Scholar] [CrossRef]
- Hussain, A.; Rizwan, R. Strategic AI adoption in SMEs: A Prescriptive Framework. arXiv 2024, arXiv:2408.11825. [Google Scholar]
- Gama, F.; Magistretti, S. Artificial intelligence in innovation management: A review of innovation capabilities and a taxonomy of AI applications. J. Prod. Innov. Manag. 2025, 42, 76–111. [Google Scholar] [CrossRef]
Dimension | Factor | Description |
---|---|---|
T1 | Compatibility | Compatibility has significant positive effect in SMEs AI adoption [94,95,96,97,98,99] |
T2 | Complexity | Complexity may be a fundamental barrier in the AI adoption [62,66,70,71,94,97] |
O2 | Cost & Financial resources | Cost is one of the mainly factors with significant negative effect [41,100,101] |
O4 | Skilled personnel | Skills shortage & AI adoption present a significant negative relationship in automation of processes, products/processes, interaction with clients, and data analytics [58,74,102,103,104,105]. |
E1 | Competitive intensity | Competitive intensity in markets presents a significant positive relationship to AI adoption [97,106,107,108] |
Region/Count. | Sector | Key Insights | Performance Indicators (KPIs) | Source |
---|---|---|---|---|
Global | Smart Manufacturing | AI-enabled smart factories. Barriers include fragmented IT, cost, and complexity. | Up to 30% increase in operational efficiency | Wan et al. (2020) [109] |
Global | Retail & Manufac. | AI + IoT improved sales, inventory, and maintenance. | 20% sales increase, 30% downtime reduction | Haider & Faisal (2024) [110] |
Europe | Multi-sector | AI + IoT + BDA synergy boosts revenue through modular strategies. | 21% increase in high-growth likelihood | Ardito et al. (2024) [111] |
Europe | General SMEs | Digital and innovation capabilities strongly influence AI success. | 52% higher AI adoption in digitally mature firms | Arroyabe et al. (2024) [27] |
Europe | Industrial | Lack of digital maturity and alignment blocks adoption. | 90% without AI apps; qualitative challenge frequency | Oldemeyer et al. (2024) [112] |
Europe | General SMEs | AI/robotics adoption enhances multi-type innovation. | 6.9% AI adoption, 8.3% robotics adoption | Segarra-Blasco et al. (2025) [113] |
Nordics | Manufac-turing | Resource orchestration key; regional networks support. | Qualitative: reduced waste, lead time, improved efficiency | Peretz-Andersson et al. (2024) [114] |
Spain | General SMEs | Digital maturity and university collaboration increase uptake. | 6x AI adoption increase via R&D links; model accuracy 85.7% | Huseyn et al. (2024) [85] |
Italy | General SMEs | Most firms lack maturity; PoCs and cloud tools help. | Only 14% use AI; 50% have no investment plans | Proietti et al. (2025) [115] |
USA | Sustaina-bility | Gen-AI supports carbon-neutral innovation | 82.8% variance in carbon-neutral performance explained | Shaik et al. (2024) [3] |
China | Finance | Internet banks outperform traditional ones for digital SMEs. | Improved loan access, no specific KPIs reported | Zhang et al. (2022) [116] |
India | Multi-sector (E-SMEs) | Digital readiness linked to strategic agility | Strong correlation with product and operational gains | Pingali et al. (2023) [5] |
Saudi Arabia | Retail | Cultural and leadership factors drive success. | Operational + economic outcomes (no % reported) | Badghish & Soomro (2024) [97] |
Africa | Finance Services | AI-fintech enhances inclusion and cuts costs. | Increased loan access, reduced operating costs | Omokhoa et al. (2024) [117] |
Pakistan | Multi-sector | Organizational factors outweigh financial ones in adoption. | Economic (), Social (), Environmental () | Soomro et al. (2025) [118] |
Ghana | General SMEs | AI in marketing drives internal and growth outcomes. | 61% variance in learning, 58% in internal process performance | Abrokwah-Larbi et al. (2024) [100] |
Challenge | Core Challenge | Strategic Solution | TOE Dimension |
---|---|---|---|
1. Knowledge and Expertise Gaps | Lack of internal technical capabilities to manage AI | Internal training, low-code tools, external partnerships | Technological/Organizational |
2. Scalability Constraints | High risk and complexity of scaling AI solutions | Modular deployment, agile methods, cloud platforms | Technological |
3. Financial Limitations | High upfront and ongoing costs for AI systems | Flexible financing, pilot testing, government incentives | Organizational |
4. Data Deficiency | Insufficient, unstructured, or poor-quality data | Data governance, real-time collection, cleaning, API integration | Technological |
5. Infrastructure and Integration Issues | Incompatible legacy systems and lack of AI-ready infrastructure | Cloud-based tools, APIs, AI computation centers | Technological |
6. Cultural Resistance to Change | Organizational fear, inertia, and lack of innovation mindset | Inclusive leadership, innovation culture, pilot-driven learning | Organizational |
7. Human–AI Productivity Misalignment | Anxiety about automation and job displacement | Augmentation approaches (co-intelligence, centaur models) | Organizational |
8. Weak Public–Private Collaboration | Limited access to shared knowledge, funding, or platforms | Innovation ecosystems, public–private partnerships, shared spaces | Environmental |
9. Limited Access to advanced AI such as Gen-AI | Barriers to using state-of-the-art AI such as Gen-AI | Plug-and-play Gen-AI tools, augmentation use cases, strategic embedding, toward open-weight LLM adoption | Technological/Organizational |
10. Lack of Responsible AI Governance | Absence of ethical practices and transparency in AI deployment | Lightweight governance, transparency, stakeholder engagement, data protection | Organizational/Environmental |
TOE Dimension | Challenge | Key Actionable Solutions |
---|---|---|
Technological | Lack of data quality and access | Data governance, real-time data collection, cleaning, integration with APIs |
Infrastructure and system misalignment | Cloud platforms, modular architectures, public–private AI centers | |
Scalability and complexity of AI tools | Modular deployment, agile methodology, scalable cloud services | |
Generative AI underutilization and lack of strategic alignment | Low-code Gen-AI tools, toward open-weight LLM adoption augmentation use cases, phased integration, innovation alignment | |
Organizational | Skills shortage and knowledge gap | Internal training, external partnerships, accessible platforms, open innovation |
Cultural resistance and lack of innovation mindset | Pilot projects, transparent leadership, learning environments, inclusive change management | |
Lack of structured methodology for AI adoption | Six-phase roadmap/methodology: assess, define strategy, select tools, pilot, train, monitor | |
Financial resource constraints | Flexible financing, pilot-based risk management, public incentives, cloud services | |
Human–AI misalignment in productivity goals | ||
Augmentation-focused adoption, co-intelligence models, upskilling, role redefinition | ||
Limited responsible AI governance and ethical practices | Lightweight AI ethics protocols, transparency, employee awareness, stakeholder inclusion | |
Environmental | Limited public–private collaboration and policy support | Innovation ecosystems, collaborative hubs, AI grants, capacity-building programs |
Phase | Key Activities |
---|---|
1. Current State Assessment | Evaluate technological infrastructure, workforce competencies, leadership alignment, digital maturity, and financial capacity. Identify strengths, weaknesses, and readiness for AI adoption. |
2. Strategic Objectives | Define clear, measurable business goals for AI adoption, ensuring alignment with overarching business priorities such as efficiency, customer experience, and data-driven decision-making. |
3. AI Solution Selection | Choose appropriate AI tools and deployment models (custom vs. off-the-shelf; cloud-based vs. on-premise) based on business needs, available resources, and scalability requirements. |
4. Pilot Project Implementation | Conduct small-scale pilot projects to validate AI solutions, assess feasibility, and build internal trust before full-scale implementation. |
5. Training and Change Management | Provide ongoing employee training to enhance AI-related skills. Implement change management practices to address resistance, align roles, and emphasize AI’s role in augmenting human capabilities. |
6. Measurement and Scaling | Establish key performance indicators (KPIs) to evaluate AI impact. Use insights from pilot phases to refine strategies and scale AI solutions across broader organizational functions. |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Sánchez, E.; Calderón, R.; Herrera, F. Artificial Intelligence Adoption in SMEs: Survey Based on TOE–DOI Framework, Primary Methodology and Challenges. Appl. Sci. 2025, 15, 6465. https://doi.org/10.3390/app15126465
Sánchez E, Calderón R, Herrera F. Artificial Intelligence Adoption in SMEs: Survey Based on TOE–DOI Framework, Primary Methodology and Challenges. Applied Sciences. 2025; 15(12):6465. https://doi.org/10.3390/app15126465
Chicago/Turabian StyleSánchez, Esther, Reyes Calderón, and Francisco Herrera. 2025. "Artificial Intelligence Adoption in SMEs: Survey Based on TOE–DOI Framework, Primary Methodology and Challenges" Applied Sciences 15, no. 12: 6465. https://doi.org/10.3390/app15126465
APA StyleSánchez, E., Calderón, R., & Herrera, F. (2025). Artificial Intelligence Adoption in SMEs: Survey Based on TOE–DOI Framework, Primary Methodology and Challenges. Applied Sciences, 15(12), 6465. https://doi.org/10.3390/app15126465