CA Sri Lanka proposes AI-powered tax compliance system for 2027 Budget

The Institute of Chartered Accountants of Sri Lanka (CA Sri Lanka) has proposed an AI-assisted tax compliance system for the 2027 Budget, arguing that the Government should improve enforcement, widen the tax base and reduce revenue leakages before considering further increases in tax rates.

Presenting its Budget proposals at a recent press conference, the Institute said Sri Lanka needed a more efficient and equitable approach to raising revenue, with greater emphasis on digitalisation, economic formalisation and voluntary compliance.

The proposals were submitted in September to Finance Ministry Secretary Dr Harshana Suriyapperuma.

CA Sri Lanka said the country remained heavily dependent on indirect taxes, placing a disproportionate burden on consumers while encouraging sections of the economy to remain outside the formal tax system.

The Institute proposed a clear order of priority for increasing revenue. It said the Government should first improve compliance under existing tax structures, identify untaxed economic activity, reduce leakages across State institutions, expand the tax base through better use of data and review existing tax exemptions.

Only after those avenues had been exhausted should increases in statutory tax rates be considered, it said.

A key proposal is the establishment of an integrated digital tax ecosystem linking information held by different State agencies.

CA Sri Lanka has recommended the creation of a National Taxpayer Information Network and a Universal Business Identification Number, which could connect information from Sri Lanka Customs, VAT records, company ownership registries, financial accounts, property records and other public databases.

The Institute has proposed that these systems, together with mandatory electronic invoicing, be developed under the broader World Bank-supported digital transformation programme.

CA Sri Lanka said the system could eventually create a unified digital economic identity, helping tax authorities identify economic activity more effectively while also simplifying the delivery of public services.

It has also called for the Inland Revenue Department and other tax authorities to move away from largely volume-based enforcement towards a Compliance Risk Management framework.

Under such a system, taxpayers would be classified according to their level of compliance risk, allowing authorities to concentrate investigative and enforcement resources on higher-risk cases.

CA Sri Lanka has proposed an AI-assisted Tax Compliance Risk Engine to support this process.

The system would analyse available data to identify potentially high-risk transactions and entities, including suspected shell companies, unexplained wealth, circular transactions and possible VAT fraud.

However, the Institute stressed that artificial intelligence should be used only as a tool for identifying and prioritising risks.

Professional judgement, due process and existing legal safeguards should continue to determine whether further investigations or enforcement action are warranted, it said.

CA Sri Lanka said the wider package was intended to increase government revenue without repeatedly placing additional pressure on taxpayers and businesses that were already complying with the law.

Caption - CEO Lakmali Priyangika, Alternate Chairman Prasad Dasanayaka, CA Sri Lanka President Tishan Subasinghe Chairman of the Faculty of Taxatiorn Saman Sri Lal, and Faculty member Athula Ranaweera.