Algorithmic TaxTech is shifting power in global wealth chains toward data-and-code owners, enabling firms to optimize legal affordances previously hidden by opacity. The change creates new infrastructural hubs that challenge traditional regulator–client–supplier relationships and could reshape who benefits from international tax rules.

Main Finding

The rise of TaxTech — especially blockchains, generative AI, and algorithmic scenario-planning — is reconfiguring global wealth chains (GWCs). Rather than only enabling secrecy, these technologies shift optimization toward algorithmic management of legal affordances, concentrating infrastructural power in suppliers (notably the Big Four) who control data, code, and platforms. The result is a move from simple information-asymmetry “cat-and-mouse” dynamics to dependencies on proprietary algorithmic infrastructures that shape tax governance, compliance, and avoidance strategies.

Key Points

• Conceptual shift: TaxTech changes the dominant mode of optimization in GWCs from hiding information to algorithmically exploiting lawful opportunities and managing compliance risk.
• Infrastructural power: Developers and platform-owners (lead suppliers) gain new governance power by controlling the technologies through which tax reporting, planning, and verification are operationalized.
• Three technology exemplars:
  • Blockchains (e.g., EY’s TaxGrid): promise near-real-time verifiability of transactions, reducing classical arbitrage/fraud (e.g., Cum‑Ex) but making both taxpayers and tax authorities dependent on platform providers and legal harmonization.
  • Generative AI / NLP: can automate routine compliance, dispute preparation, and extract structured insights from heterogeneous, unstructured tax data; accelerates outsourcing and creates high demand for hybrid tax+data skills, favoring large firms that can hire and license models.
  • Algorithmic scenario planning (described as a third strand): uses simulation/optimization to search complex legal affordances and craft tax-minimizing structures, shifting opacity from hidden transactions to opaque machine reasoning and model outputs.
• Power and dependency dynamics: Technologies can reduce some traditional information asymmetries (helping regulators) while simultaneously increasing supplier leverage: control over proprietary models, data access, and integration creates vendor lock-in and new forms of captured GWCs.
• Promotional bias & performance gaps: Evidence comes largely from Big Four sources (webcasts, marketing); many claims may be performative, and implementations may be “quasi‑AI” or limited in practice.

Data & Methods

• Multi-phase qualitative approach focused on the emerging TaxTech field (2017–2022):
  • Phase 1: Systematic review of Big Four publications, websites, and digital strategy materials to map products, collaborations, and imaginaries.
  • Phase 2: Collection and transcription of ~15 public webcasts/online events hosted by Big Four and tech partners; used to identify actors and claims.
  • Phase 3: Follow-up interviews (a small number) with tax professionals; supplemental reliance on suppliers’ marketing and product descriptions for up-to-date status.
• Analytical frame: Global Wealth Chains (GWC) literature to interpret power relations among suppliers, clients, and regulators.
• Limitations acknowledged by authors:
  • Heavy reliance on supplier-produced materials (marketing bias).
  • Emerging technologies may be over-sold; many offerings are promotional or incremental rather than genuinely novel.
  • Small number of interviews; no large-scale quantitative validation of realized impacts.

Implications for AI Economics

• Market structure and rents:
  • Ownership of models, proprietary datasets, and integrated platforms can create durable market power for large accounting/tech firms; rents may shift from taxpayers/MNEs toward platform suppliers.
  • Vendor lock-in and network effects (e.g., standardized blockchain webs) risk entrenching dominant suppliers and raising switching costs.
• Information asymmetry and regulatory capacity:
  • TaxTech can both reduce verifiability gaps (benefiting regulators) and create a new asymmetry: regulators may depend on private algorithms and data-sharing agreements, weakening independent oversight.
  • The opacity of algorithmic reasoning introduces a new enforcement challenge (algorithmic opacity vs. previous transactional opacity).
• Labor and skill composition:
  • High demand for hybrid professionals (tax expertise + data science/coding) will reallocate labor and raise barriers to entry for smaller advisory firms; outsourcing/co-sourcing of tax functions likely to increase.
• Efficiency vs distributional consequences:
  • Possible macro benefits: lower compliance costs, reduced large-scale fraud, more accurate tax bases.
  • Possible distributional harms: concentration of governance power, greater capture of value by platform owners, and unequal access to optimization tools (favoring large MNEs that can pay).
• Policy and regulatory research agenda:
  • Competition policy: examine market concentration among advisory/TaxTech suppliers and the implications of platform-led GWCs.
  • Model and data governance: require model auditability, provenance of training data, and standards for explainability for tax-related algorithms.
  • Interoperability & public infrastructure: evaluate public or multi-stakeholder blockchain/web standards to avoid single‑vendor capture while preserving verifiability benefits.
  • Transparency rules: design reporting standards for algorithmic tax-planning outputs and automated advice to enable regulatory review.
• Empirical research directions:
  • Measure realized effects on tax revenues and avoidance after adoption of TaxTech (natural experiments, diff‑in‑diff where pilots occur).
  • Quantify changes in market concentration and rents captured by suppliers post‑adoption.
  • Study labor-market impacts: demand for tax+AI skills, outsourcing rates, wage premia.
  • Analyze how algorithmic scenario-planning changes observable firm behavior (entity structures, profit location) compared to pre‑algorithmic planning.

Summary takeaway: TaxTech has the potential to both strengthen tax verifiability and concentrate new forms of infrastructural power in private suppliers. For AI economics, this creates a rich set of questions about market power, regulation of algorithmic infrastructures, labor reallocation, and the net public finance consequences of technological adoption.