Accountants in Isabela use AI largely for routine tasks, improving efficiency and accuracy, but advanced applications are scarce; weak workforce skills and limited infrastructure are the main barriers to broader adoption, pointing to a need for targeted training and investment.

Main Finding

Accounting firms in Santiago City and Cauayan City (Isabela, Philippines) are in early stages of AI adoption. Practitioners report strong foundational analytical skills but primarily use AI for repetitive, routine tasks (data entry, reconciliation, basic automation). Higher‑level uses (advanced analytics, sustained AI governance, strategic advisory) remain limited. Perceived benefits include improved efficiency, accuracy, compliance, and decision quality, but adoption is constrained by workforce skill gaps and inadequate supporting infrastructure.

Key Points

• Sample and context

  • Survey of 28 accounting practitioners across 17 firms (turnout 54.8%) in Santiago City and Cauayan City, Isabela.
  • Most firms are sole proprietorships serving mostly local clients; ~46% of firms report >100 clients.
  • Adoption is recent: ChatGPT used by 61% of firms, Microsoft CoPilot by 25%; 7 firms reported AI not applicable.

• Adoption profile & roles

  • Rogers adopter categories: Innovators 28.6%, Early Adopters 25.0%, Early Majority 32.1%, Laggards 14.3% (no Late Majority respondents).
  • Self‑reported roles in AI processes: Analyzers 46.4%, Identifiers 28.6%, Explainers 17.9%, Sustainers 7.1% — indicating emphasis on interpretation/insight over maintenance/governance.

• Competencies & gaps

  • High prevalence of critical thinking and analytical skills (75%).
  • Data analytics competency reported by 64.3%.
  • Lower prevalence of direct AI tool knowledge (42.9%) and ethical proficiency in AI use (42.9%).
  • 14.3% reported none of the listed competencies.

• Uses & benefits

  • AI predominantly applied to routine, repeatable accounting tasks (OCR, data extraction, reconciliation).
  • Reported benefits: operational efficiency, accuracy, compliance improvement, and enhanced decision support.
  • Auditing practitioners (business owners) reported higher AI use frequency than other roles.

• Barriers

  • Main barriers: lack of workforce skills, limited infrastructure, and limited long‑term institutional support or governance capacity.

Data & Methods

• Design
  • Comparative research design using descriptive and inferential statistics, supported by thematic analysis of qualitative responses.
• Sample
  • N = 28 practitioners from 17 accounting firms in two cities in Isabela, Philippines.
• Instruments & procedures
  • Self‑developed questionnaire (four parts: profile & adopter role; competencies; extent of utilization across bookkeeping/financial reporting/auditing/taxation; perceived benefits).
  • Pen‑and‑paper surveys with follow‑up interviews for validation.
  • Ethical procedures: informed consent, confidentiality, minimal disruption.
• Quantitative analysis
  • Descriptive statistics (frequencies, means, medians, variance).
  • Normality assumption violated → nonparametric tests: Mann‑Whitney U and Kruskal‑Wallis (alpha = 0.05).
• Qualitative analysis
  • Thematic analysis (iterative coding and constant comparison) to synthesize perceived benefits and narratives.
• Limitations (reported or implied)
  • Small, localized sample; self‑reported measures; cross‑sectional design; limited managerial representation (no managers in sample).

Implications for AI Economics

• Productivity and task reallocation
  • Current adoption mostly automates routine tasks, implying near‑term productivity gains primarily via time savings and error reduction. Broader economic gains (higher value tasks, advisory revenue) are potential but unrealized without skill upgrading.
• Labor demand and skill premium
  • Demand will shift toward analysts and AI‑literate accountants (data analytics, model interpretation, ethical oversight). A scarcity of AI skills suggests upward pressure on wages for those competencies or a need for upskilling programs to avoid displacement or underemployment.
• Investment patterns & firm heterogeneity
  • Small local firms adopt lightweight, off‑the‑shelf GenAI tools (e.g., ChatGPT) rather than integrated enterprise systems, reflecting lower capital expenditure but faster diffusion of accessible AI. Economic models should account for heterogeneous adoption paths (toolbox vs. platform investment).
• Diffusion dynamics
  • Adopter distribution skews toward innovators and early adopters, suggesting rapid local diffusion among engaged practitioners but potential lag among resource‑constrained laggards. Policy or network effects (peer demonstration, owner leadership) can accelerate uptake.
• Complementarities & infrastructure
  • Limited infrastructure and governance (few sustainers) imply underinvestment in complementary assets (data governance, integration, secure IT). Without these investments, firms may realize only partial benefits, reducing the social return on AI adoption.
• Policy and market interventions
  • Targeted training subsidies, public–private partnerships, and low‑cost infrastructure support (secure cloud/managed services, compliance tools) could raise adoption quality and aggregate economic benefits.
• Implications for auditing/taxation markets
  • Higher uptake in auditing suggests those submarkets may evolve faster (larger scope for anomaly detection, full‑population testing). Tax compliance and fraud detection could create public finance gains if scaled.
• Research & measurement
  • For macroeconomic assessments, small‑firm, localized studies like this underline the need to measure both intensity (how tools are used) and capability (skills/infrastructure) — mere adoption counts overstate economic impact if advanced uses remain limited.

Suggestions for researchers and policymakers - Prioritize longitudinal studies to track whether routine automation evolves into higher‑order AI use (advisory, forecasting). - Design upskilling programs focused on AI tool literacy, data analytics, and AI ethics for accountants in SMEs. - Support affordable infrastructure and governance frameworks that enable secure, integrated AI deployment in small firms.

If you want, I can (a) expand this into a one‑page policy brief aimed at regional policymakers, (b) extract recommended training curriculum topics for accountants, or (c) transform this into slides for a presentation.