This research uses an interpretive case study strategy to investigate how big data affects tax audits in Indonesia, both with regard to tax audit management and policy, and to tax auditors' individual audit assignments. The study reveals that the impact of big data on tax audit exists in two aspects. First, at audit policy level, big data is used as part of risk analysis in order to determine which taxpayers should be audited. Second, at the individual tax audit assignment level, tax auditors must utilise big data in order to acquire and analyse data from taxpayers and other related parties. Big data has the following characteristics: it involves huge volumes of information, it is generated at a high velocity, it includes a wide array of data types, and it contains high uncertainty. Big data can be analysed in order to reinforce the results gained from risk engines as a part of a compliance risk management system at the audit policy level. Meanwhile, at the individual tax audit assignment level, empirical evidence shows that tax auditors may deal with: (1) large volumes of data (hundreds of millions of records) that originated from previous fiscal years (historical records); (2) variations in the format and sources of data acquired from taxpayers which, to some extent, may be giving an auditor the authority to request data in a format that suits their analytical tools-with an inherent risk that the data can only be acquired in its native format; (3) data veracity that requires the tax auditors to review data sources because the adopted data analysis techniques are determined by the validity of data under audit. The main benefit expected to be gained from the implementation of big data analytics in respect of tax audits is the provision of valid and reliable information that evidences that taxpayers are compliant with tax laws.
