Zhou Keqing, Chen Fengjin and Yang Yi analyse in the September 2026 issue of Taxation Research how the tax system should accommodate AI Plus. Their assessment is that the current system still follows traditional industrial-economy thinking and lags in three visible ways: R&D incentive rules do not fit inputs dominated by intangible assets, asset tax rules do not keep pace with changing asset forms, and VAT rules do not keep pace with digital transaction services. Borrowing from international practice, they recommend widening the boundary of R&D tax incentives, filling in rules for the amortisation, transfer and investment of data resources, and optimising VAT so that it covers digital transaction services.
News Detail
How the tax system can empower AI Plus: status, experience and reform paths
In the September 2026 issue of Taxation Research, Zhou Keqing, Chen Fengjin and Yang Yi argue that AI has become a core engine for new quality productive forces and industrial upgrading, yet a tax system built on traditional industrial-economy logic still falls short of supporting it. They identify three mismatches: R&D incentive rules that do not fit input structures driven by intangible assets, asset tax rules that lag changes in asset forms, and VAT rules that lag the growth of digital transaction services. Drawing on international practice, they recommend widening the boundary of R&D tax incentives, completing rules for the amortisation, transfer and investment of data resources, and optimising VAT so it covers digital transaction services.
2026-11-05