As India’s global capability centres move from back-office support to hubs for product, engineering and analytics, their tax models are struggling to keep pace. Transfer pricing arrangements set years ago often no longer reflect where decisions are made or which risks are managed. Tax teams, meanwhile, are pulled in only after contracts are signed and invoices are paid, leaving them to reconstruct compliance at month-end.
Alok Jain, Head of Tax at Embark, argues that the fix begins with sequencing. Tax must sit closer to the transaction, inform location and funding choices early, and be designed into the operating model from day one. In this conversation with CIO&Leader, he explains what “tax by design” means in practice, why a “strategic” label alone does not change transfer pricing outcomes, where automation and AI genuinely help, how to separate sanctioned AI from shadow AI, and how much autonomy India teams should have from HQ—his conclusion: the real advantage is not a lower tax bill, but the absence of friction.
CIO&Leader: Tax needs to be designed into the GCC operating model from the outset, rather than added later. What does “tax by design” actually look like in practice? Which setup decisions have the biggest downstream tax consequences?
Alok Jain: For a GCC, “tax by design” means understanding what activities will sit in India and how the India entity is expected to operate within the wider organisation. This directly affects the transfer pricing model. A GCC performing defined support services under a limited-risk model is very different from one expected to make important decisions, develop IP or take on significant business risks. The intercompany arrangements and remuneration model should therefore reflect the role the GCC is expected to play.
Beyond transfer pricing, decisions regarding the legal entity, location, available incentives, funding, and repatriation can have fairly long-term consequences. Systems matter as well. If the Enterprise Resource Planning (ERP) system and processes are not designed to capture the correct tax information, some of these requirements may be managed each month manually.
In that sense, tax by design is less about treating tax as a separate consideration and more about having it considered alongside the wider business and operating model. This can provide greater clarity around the tax implications of decisions while the GCC structure is still being developed.
CIO&Leader: Where do you see the biggest gaps in existing tax operating models as GCCs scale from support function to a more strategic, complex mandate?
Alok Jain: One issue I see often is that the GCC evolves, but the tax model still reflects what it was originally set up to do. A centre may start with IT support and, a few years later, have teams working on product, engineering, or analytics, with much more decision-making happening in India. The transfer pricing model and intercompany arrangements do not always move at the same pace.
That does not automatically mean the tax outcome has to change. What matters is what has actually changed on the ground, including the work being done, the risks being managed and where decisions are being made.
Tax can also get involved later in the process, often after the contract is signed or the transaction has taken place, as the GCC grows and becomes more complex, bringing that perspective earlier in the process.
The India tax model also needs to fit within the parent organisation’s overall tax framework. There should be consistency in policy, controls, and reporting, with enough flexibility to address India-specific requirements locally.
CIO&Leader: Industry bodies are pushing for wider GST input tax credit on employee-related costs: food, catering, health insurance, vehicle leasing. How much of a cost burden is this for people-intensive GCCs today, and what would change if these restrictions were eased?
Alok Jain: For a people-intensive GCC, the context has changed significantly. Health insurance, food, transportation, and similar benefits are no longer peripheral employee benefits. They are part of the employee proposition, particularly when GCCs compete for specialised technology, engineering, and finance talent.
The GST law has historically restricted the availability of credit for several of these expenses because they were seen as having an element of personal consumption or employee benefit. But for a large GCC today, many of these are simply costs of running the workplace and attracting people.
When GST credit is unavailable, the tax becomes a cost. It would be difficult to put a generic percentage on the impact because every GCC has a different benefits structure, but at scale the amount can be meaningful.
If these credit restrictions were eased, it could reduce some of that embedded cost and bring the tax treatment closer to how modern, people-intensive businesses incur these expenses.
CIO&Leader: Transfer pricing is often the most contentious tax issue for GCCs. As centres take on higher-value, more strategic work for their parent organisations, how should transfer pricing frameworks evolve? How should they keep pace?
Alok Jain: I would look at how the GCC’s role has actually evolved. Describing a centre as “strategic” or a “centre of excellence” does not, by itself, mean that the transfer pricing model needs to change.
What matters is the nature of the activities being undertaken in India. Greater decision-making, management of additional risks or involvement in IP development can all be relevant considerations. At the same time, taking on more sophisticated work does not necessarily result in a different transfer pricing outcome.
The position can also be revisited as the GCC matures. A cost-plus model that was appropriate for a 100-person centre may still be appropriate at 1,000 people, depending on the functions, assets and risks involved. Size alone does not determine the outcome.
Safe harbour and APAs can provide certainty, but the transfer pricing position ultimately needs to reflect how the GCC operates in practice.
CIO&Leader: What role can automation and AI realistically play in reducing operational tax work for GCCs, and where are teams over-relying on technology versus where is it still under-used?
Alok Jain: AI can add significant value to tax operations, although its effectiveness depends on the processes and data it is applied to. First, establish whether a process is standardised, whether the underlying data is reliable and which activities are rule-based.
Many areas of tax work can be addressed through clear rules or straightforward automation, while AI is more relevant for unstructured information, exceptions and areas that require interpretation or judgment.
Applying AI on top of a fragmented process may make it faster, but it does not necessarily make it better. Get the process and data right first, automate what is predictable and bring AI in where it genuinely adds something.
CIO&Leader: In the “Operate” phase specifically, what do you see companies consistently getting wrong on tax, and what does getting it right look like?
Alok Jain: One thing I have seen consistently is that tax is treated as a downstream process. A transaction occurs during the month, and the tax assessment is made later, when someone starts preparing for withholding tax or GST compliance. By then, the invoice may already have been recorded or even paid.
That sequence needs to change. Tax needs to sit much closer to the transaction. When an invoice is recorded, the tax treatment should be clear at that point, including the withholding position, GST treatment, credit eligibility and any required documentation.
This also changes what the tax team spends its time doing. Instead of finding and fixing issues at month-end, the team can focus on exceptions and matters that require judgment.
A good Operate model should make compliance an output of transactions occurring during the month, rather than a separate exercise to reconstruct them afterwards.
CIO&Leader: You describe stronger controls, visibility and scalable processes as a competitive advantage. Can you give a concrete example of how better tax visibility has changed an operating decision for a GCC? Can you also give an example of how it has changed an investment decision?
Alok Jain: Tax visibility is most useful when information reaches management early enough to inform a decision. P&L and cash positions are typically available relatively quickly, whereas tax information often comes later in the process.
Take GST refunds. If an export-heavy GCC can see that GST credit is consistently building up rather than being recovered, that is not just something for the tax team to explain at quarter-end. It affects working capital and can change how the refund process is managed.
The same applies when a GCC is planning a significant expansion in India. Tax incentives, transfer pricing implications, and the cost of funding and repatriation should be part of the investment case as it is being built. Finding these out after the location and structure have been decided is too late.
To me, that is what tax visibility means. The information needs to reach management early enough to be useful.
CIO&Leader: How does Embark view the trade-off between centralising tax decision-making at the global HQ level and giving the India GCC greater autonomy over its own tax function?
Alok Jain: At Embark, we do not think there is one answer for every GCC. A 200-person centre with a small finance team should operate differently from a mature GCC with several thousand people and a strong India tax team.
Much depends on how the GCC is aligned with the parent organisation. The group tax framework, risk positions and matters requiring global approval should be clearly defined. Within those boundaries, the India team should have sufficient autonomy to handle local tax matters without sending routine decisions back to HQ. That autonomy can increase as the GCC and the local team mature. Material cross-border positions, changes to the transfer pricing model or matters affecting the wider group would still need global alignment.
It is less about giving India independence from HQ and more about giving the local team enough ownership to operate effectively within an agreed framework.
CIO&Leader: With India’s GCC ecosystem maturing rapidly, how are state-level incentives, SEZ/GIFT City structures, and evolving compliance requirements reshaping where and how companies set up new centres?
Alok Jain: Talent and the operating model still come first when evaluating a GCC location. An attractive incentive package cannot compensate for a location that does not meet the business’ talent and capability requirements.
That said, incentives have become a meaningful part of the location decision, with states competing for GCC investment through employment, skilling, capital, and other incentives. Special Economic Zones (SEZs) can remain relevant for some export-oriented operations, while Gujarat International Finance Tec-City (GIFT City) serves a more specific set of financial services activities.
The starting point is to identify where the GCC can succeed, and then assess how the available tax and incentive framework can strengthen the economics.
CIO&Leader: Five years from now, do you expect tax to be seen as a cost centre for GCCs, or as a strategic differentiator companies actively compete on? What needs to happen for the latter to become the norm?
Alok Jain: Tax becomes a differentiator when the GCC can take on more complex work without creating additional risk or uncertainty for the parent. As more responsibility moves to India, the tax questions naturally get harder. Transfer pricing, IP, cross-border arrangements and employee mobility start becoming more relevant. If every change in the GCC mandate leads to a lengthy tax review or creates uncertainty at HQ, it can slow the business down.
A strong tax function should make that transition easier. The basic framework should already be clear, routine compliance should not consume the team, and tax should get involved early enough to work through a change before it becomes a problem. The advantage is not a lower tax bill. It is that the tax is not the reason the GCC cannot take on the next piece of work.