Digital VAT in Sri Lanka: Why platform liability and reverse charge are essential to avoid costing households and Treasury

Executive summary

Sri Lanka’s VAT on cross-border digital services, in force since July 2026, is a necessary and overdue reform, but its design is incomplete. It relies only on registration by non-resident suppliers and fully carves out supplies to VAT-registered businesses. Without platform liability and a B2B reverse charge, it will likely under-collect and place the entire new burden on households.

Key points

Background: The reform was IMF-backed, with a modest expected yield (0.04-0.08% of GDP). Its stronger case is equal treatment of local and foreign suppliers. It was delayed twice before Act No. 14 of 2026 confirmed the July start.

Weak enforcement: The Inland Revenue Department would have to chase foreign vendors with no local presence. Large brands may register, but smaller providers likely won’t.

B2B carve-out: (Supplies made to local VAT registrants): Reverse charge would raise real revenue from banks, financial institutions, life insurance companies and other VAT registered companies who are prevented from claiming full VAT input credit, which are exempt from invoice-credit VAT and therefore cannot claim input tax credits under that mechanism, while also creating an audit trail. The carve-out forgoes this, loses visibility, and leaves local providers at a disadvantage.

Equity: Individuals pay digital VAT while businesses buying the same services pay nothing, which undercuts the aim of base broadening and any future cut to the 18% rate.

Global practice: Simplified registration for B2C supplies and reverse charge for B2B supplies are widely used internationally. Platform liability is also increasingly adopted, including in the EU, UK, Australia, Canada and Singapore, but its scope varies. OECD guidance presents platform liability as an effective policy option. 

Recommendation: Keep simplified registration for B2C supplies, add platform liability for marketplaces, app stores, and replace the B2B carve-out with reverse charge. Support these with clear location-indicator guidance, proportionate penalties and data-sharing with financial institutions.

Completing the framework with platform liability and reverse charge

Sri Lanka’s decision to tax cross-border digital services represents a necessary and overdue modernisation of the Value Added Tax system. It responds to the reality that value is increasingly created and consumed online, often without any physical presence in the country of consumption. 

Local providers have long charged VAT on comparable services while many non-resident suppliers have not. The reform aims to correct that imbalance, broaden the tax base, and support the eventual possibility of reducing the high 18% standard rate.

 Yet the framework that has been enacted relies almost exclusively on vendor registration by non-resident suppliers and contains a complete carve-out for supplies to VAT-registered businesses. This design is incomplete. Without the addition of platform/marketplace liability and a reverse charge mechanism for business-to-business supplies, the regime risks delivering only a fraction of its potential revenue while placing the entire incremental tax burden on individuals and households. Policymakers still have the opportunity to strengthen the legislation and the accompanying administrative rules so that digital VAT becomes a genuine instrument of base broadening rather than a limited and largely symbolic measure.

A record of ambition tempered by repeated delay

The intellectual and policy origins of digital VAT in Sri Lanka are well documented. The International Monetary Fund’s June 2024 staff report identified the taxation of digital services as one element of a broader package intended to raise revenue and improve neutrality. The expected yield was modest, approximately 0.04% of GDP in the first full year, with some estimates ranging up to 0.08%. The stronger argument was never the absolute size of the revenue but the principle of equal treatment: a Sri Lankan consumer purchasing streaming, software or cloud services from a foreign provider should face the same VAT treatment as one purchasing from a local supplier.

Legislative progress followed a familiar pattern of announcement, enactment and postponement, further prolonged by a Supreme Court challenge that delayed the regime’s finalisation and exacted a real opportunity cost in foregone revenue. The Value Added Tax (Amendment) Act No. 4 of 2025 was certified on 11 April 2025. Gazette Notification 2443/30 of 1 July 2025 set an original effective date of 1 October 2025. Cabinet subsequently deferred implementation to 1 April 2026. A further postponement moved the start to 1 July 2026. The Value Added Tax (Amendment) Act No. 14 of 2026, certified on 30 June and published in gazette on 3 July 2026, finally placed the regime on a firm statutory footing and confirmed the July 2026 commencement.  

Although Sri Lanka’s digital-services VAT provisions came into effect on 1 July 2026, with section 25L principally governing registration and section 25N providing that VAT is not charged where the recipient is VAT-registered, the regime remains non-operational in practice because the prescribed registration form has yet to be made available, preventing non-resident digital service providers from completing the registration process.

The mechanics that emerged are conventional in outline but incomplete in substance.

Non-resident suppliers of digital services through electronic platforms must register once their supplies to persons in Sri Lanka exceed Rs. 60 million in any twelve-month period or Rs. 15 million in a quarter. A recipient is treated as located in Sri Lanka if at least two of the following indicators are present: a billing or residential address in Sri Lanka, payment through a Sri Lankan bank or financial institution, a payment instrument issued in Sri Lanka, or a Sri Lankan IP address. Registration, return filing and payment are electronic; payment may be made in Rupees or an approved foreign currency, either directly or through a representative. Supplies to VAT-registered resident businesses are entirely carved out. No obligation is placed on platforms, marketplaces or app stores to collect or remit VAT on the underlying supplies they facilitate.

Digital VAT: Real revenue, not a compliance checkbox

This sequence of delays and the final design choices invite a sober assessment. Digital VAT should not be reduced to a compliance checkbox required by an external program. It should raise meaningful revenue, follow the proven architecture used by successful jurisdictions, and contribute to the larger project of base expansion that can eventually support a lower standard rate. Features that work elsewhere should not be discarded because of lobbying pressure, administrative convenience or incomplete understanding of implementation options. More than one hundred countries already collect VAT on non-resident digital services. The tax base belongs to the country of consumption. Sri Lanka is entitled to collect it fully.

Introducing marketplace liability and reverse charge would convert a half-measure into a credible instrument of fiscal policy, one capable of raising revenue, promoting fairness between local and foreign suppliers, and supporting the longer-term objective of reducing the high standard VAT rate of 18% that currently weighs on consumers. The tools exist. International experience demonstrates that they work. The remaining question is whether Sri Lanka will choose to use them

OECD recommendations and the global standard

The OECD International VAT/GST Guidelines, together with subsequent practical toolkits, provide the clearest international consensus on how to tax cross-border supplies of services and intangibles. For business-to-consumer supplies the recommended approach is to require the non-resident supplier to register and account for the tax under a simplified regime. For business-to-business supplies the preferred mechanism is reverse charge: the local business customer accounts for the VAT as if it were the supplier.

This arrangement relieves the foreign supplier of registration obligations on those particular supplies while ensuring that the tax is reported and, where the customer cannot claim full input credit, that revenue is actually collected. Where digital platforms facilitate the transactions, the OECD strongly encourages jurisdictions to impose liability on the platform itself—the “deemed supplier” or full platform liability model. Platforms already possess customer location data, control payment flows and can scale compliance far more efficiently than a tax administration attempting to locate and compel thousands of individual foreign vendors.

International practice is not uniform. Simplified vendor registration for B2C supplies and reverse charge for B2B supplies are widely used across jurisdictions. Platform liability is also increasingly adopted, including in the European Union, the United Kingdom, Australia, New Zealand, Canada and Singapore, but its scope differs. OECD guidance identifies platform liability as an effective policy option.

Australia’s extension of GST to imported digital services and the European Union’s progressive introduction of deemed-supplier rules illustrate the practical advantages of concentrating collection at the platform level. Enforcement becomes realistic, the long tail of smaller suppliers is captured, and revenue leakage is substantially reduced.

Sri Lanka’s exclusive reliance on vendor registration, the absence of platform liability, and the substitution of a full B2B carve-out for reverse charge constitute the weakest available combination of collection methods. Large, reputation-conscious global brands may eventually register. The multitude of smaller software providers, freelance platforms, content creators and niche applications almost certainly will not. The Inland Revenue Department will be left to pursue entities that have no local presence, no local bank account and little commercial incentive to comply. When the same administration already faces well-documented challenges in securing full compliance from resident taxpayers, the expectation that it can effectively police non-residents who simply ignore registration or under-declare values is unrealistic. 

Platform liability reverses the burden of enforcement: a limited number of large, visible platforms become the collection points, using systems and data they already maintain.

The revenue and equity cost of the B2B carve-out

The decision to exempt supplies to VAT-registered businesses rather than apply reverse charge carries particularly serious consequences for both revenue and fairness. 

Under a reverse charge mechanism a local VAT-registered recipient self-assesses the tax. For a fully taxable business the net cash effect is neutral because the output tax is matched by an equal input credit. Banks, financial institutions, life insurance companies and other VAT registered companies who are prevented from claiming full VAT input credit are subject to the separate VAT on financial services but are exempt from invoice-credit VAT; accordingly, they cannot claim input tax credits under the invoice-credit mechanism. The same concern may arise for other businesses whose ability to recover input tax is restricted. Those sectors are significant purchasers of foreign cloud computing, software licenses, digital advertising, data analytics and related services. Reverse charge would therefore generate genuine additional revenue from a rapidly growing category of imports. It would also create a complete audit trail that allows the tax authority to monitor the volume and nature of cross-border digital purchases.

Under the B2B carve-out, the foreign supplier does not charge VAT and the VAT-registered recipient does not self-assess it. However, the enacted framework may still require the foreign supplier to furnish prescribed information concerning such supplies to the Inland Revenue Department. The carve-out therefore does not necessarily leave these transactions entirely unreported, but it does not create the same recipient-level self-assessment and audit trail that a reverse charge would provide. This limits visibility and may make misclassification, for example, a private consumer presenting as a business is more difficult to detect.

Local digital service providers continue to charge VAT while their foreign competitors do not on B2B sales, recreating the competitive distortion the reform was intended to eliminate.

This outcome stands in direct contradiction to the stated policy objective of base broadening. Broadening the tax base cannot mean capturing only individuals, households and non VAT registrant businesses, while excluding businesses that are exempt from invoice-credit VAT, including banks, financial institutions, life insurance companies and other VAT registered companies who are prevented from claiming full VAT input credit, from the reverse-charge net. Yet that is the practical effect of the present design. 

Individuals, households and non VAT registrant businesses pay digital VAT on streaming subscriptions, applications, software and online services. Businesses purchasing comparable or identical services from non-resident suppliers face no corresponding liability. The entire additional burden of expanding the VAT base in the digital sphere therefore falls on ordinary consumers. Such an arrangement is inequitable. It is also fiscally short-sighted. Every Rupee of revenue left uncollected from B2B digital imports is a Rupee that cannot contribute to the fiscal space required for a future reduction in the standard 18% rate.

Fairness, administrative capacity and the question of influence

The present design raises uncomfortable questions of fairness. 

Domestic VAT registration thresholds have been subject to repeated adjustment; more local small and medium enterprises face compliance obligations and the associated costs. At the same time the largest global platforms face only a registration form that they may or may not complete, with no marketplace liability to enforce collection on the many underlying suppliers they host. Local providers charge VAT; foreign providers of B2B digital services do not. The competitive playing field remains tilted.

It is legitimate to ask whether the abandonment of reverse charge and the omission of platform liability reflect purely technical judgment or the successful representation of interests that prefer the current, limited arrangement. 

Reverse charge, by its ordinary operation, would have brought businesses exempt from invoice-credit VAT, including banks, financial institutions, life insurance companies and other VAT registered companies who are prevented from claiming full VAT input credit, into the collection mechanism and produced revenue that the carve-out simply forgoes. Administrative convenience is not a sufficient justification for forgoing that revenue. Other countries with larger digital markets and more complex economies have implemented the fuller OECD package. Sri Lanka does not need to invent new systems. It can adapt existing models and begin with the largest platforms. 

The alternative, proceeding with vendor registration alone and a full B2B carve-out, almost guarantees modest receipts, persistent non-compliance among smaller foreign providers, zero contribution from the B2B segment, and, within a relatively short period, another round of legislative amendments to correct what could have been designed correctly from the outset.

A practical and proven path forward

A more robust framework is available and does not require radical innovation. It would retain simplified electronic vendor registration for business-to-consumer supplies, introduce platform liability for major marketplaces, app stores, and replace the B2B carve-out with a reverse charge mechanism. 

Supporting measures, clear guidance on the application of location indicators, proportionate penalties, and structured information exchange with financial institutions would complete the architecture. These are precisely the tools that have proven effective in other jurisdictions. They align with OECD recommendations, protect the competitive position of local providers, generate revenue from purchases by businesses exempt from invoice-credit VAT, including banks, financial institutions, life insurance companies and other VAT registered companies who are prevented from claiming full VAT input credit, and ensure that base expansion is not borne solely by households.

Sri Lanka cannot afford a digital VAT regime that exists primarily to satisfy an external checklist. The digital economy is one of the few expanding tax bases available to the country. Confining collection to vendor registration while carving out the entire B2B segment is, in effect, planning for underperformance. Policymakers still possess the authority and the time to strengthen the regime before it settles into underwhelming practice. 

Introducing marketplace liability and reverse charge would convert a half-measure into a credible instrument of fiscal policy, one capable of raising revenue, promoting fairness between local and foreign suppliers, and supporting the longer-term objective of reducing the high standard VAT rate of 18% that currently weighs on consumers. The tools exist. International experience demonstrates that they work. The remaining question is whether Sri Lanka will choose to use them.

(The author was awarded Tax Practice Leader of the Year 2024 (ASPAC) by International Tax Review (ITR) and was a top-four finalist for Tax Litigation and Disputes Practice Leader of the Year (ASPAC))