What Changed in European Tax in 2026

Three things changed in Brussels in the first weeks of 2026. One of them switches off a rule that American groups had spent two years preparing for. Another turns carbon into a line item on an import invoice for the first time. The third has been quietly in force since 2022 and is still the most underused instrument available to any company moving money between Bogotá and Europe.

Pillar Two: the UTPR safe harbour for American groups

The EU Minimum Tax Directive — Council Directive (EU) 2022/2523 — brought the OECD's fifteen per cent global minimum tax into European law. For a US-parented group with a Colombian subsidiary and any European presence, it raised an uncomfortable question: could a European member state reach across and top up tax on profits earned outside Europe entirely?

Under the Undertaxed Profits Rule, the answer was yes. That is the mechanism that allows one jurisdiction to collect top-up tax on low-taxed profits arising anywhere in the group, including in jurisdictions with no connection to the collecting state.

On 5 January 2026 the OECD released what has become known as the side-by-side package. On 12 January the European Commission published a notice in the Official Journal confirming that the package applies through Article 32 of the Directive — without any amendment to the Directive itself.

The practical effect is significant. The side-by-side safe harbour switches off both the Income Inclusion Rule and the Undertaxed Profits Rule where the ultimate parent entity sits in a jurisdiction that imposes minimum taxation on both domestic and foreign income and grants a foreign tax credit for qualified domestic minimum top-up taxes. On the OECD's central record, the United States is currently the only jurisdiction that qualifies.

The safe harbour is available for fiscal years beginning on or after 1 January 2026. Two cautions belong alongside that sentence. First, availability depends on local implementation, and in many member states the rules will not be transposed into national law before the second half of 2026 or even 2027. Second, this is a safe harbour, not a repeal: the underlying obligations, the data collection, and the qualified domestic minimum top-up taxes in each jurisdiction remain exactly where they were.

For a US group that spent 2024 and 2025 building Pillar Two capability, the correct reading is not that the work was wasted. It is that the exposure has narrowed sharply, and the compliance calendar for the next two years is now a question of which member states have transposed what, and when.

CBAM: from report to cost

The Carbon Border Adjustment Mechanism entered its definitive phase on 1 January 2026. For two years it had been a reporting exercise. It is now a financial one.

CBAM covers six sectors: cement, iron and steel, aluminium, fertilisers, electricity and hydrogen. Importers bringing more than fifty tonnes of in-scope goods into the European Union in a year must hold authorised CBAM declarant status — and only an authorised declarant may import above that threshold. Below fifty tonnes, the obligations do not bite.

The money follows on a delay. Certificate sales open on 1 February 2027, and the first surrender of certificates takes place in 2027, covering goods imported during 2026. Pricing for 2026 imports reflects the quarterly average auction price of EU emissions allowances; from 2027 the reference shifts to a weekly average.

This matters on both sides of the Atlantic, and for different reasons. A Colombian producer selling steel, aluminium or cement into Europe will find that the carbon intensity of its production is now a commercial variable, priced by its customer's compliance cost rather than by any Colombian regulation. A European buyer will start asking for emissions data with the same insistence it applies to certificates of origin.

The uncomfortable part is that the data has to come from somewhere. Embedded emissions are calculated at installation level. A producer that cannot substantiate its figures will find its goods priced on default values, and default values are not designed to flatter.

The Colombia–France tax treaty almost nobody uses properly

The double taxation convention between Colombia and France was signed in June 2015. It then waited. The Colombian Congress approved it in October 2020, the Constitutional Court upheld it on 9 December 2021, and it entered into force on 1 January 2022. Under Article 30, its provisions took effect from 1 January 2023 for taxes withheld at source.

Built on the OECD model, it reduces withholding rates on dividends, interest and royalties between the two countries. In a region where cross-border payments routinely attract full statutory withholding because nobody checked whether a treaty applied, that is not a technicality. It is a recurring, quantifiable leak.

We see the same pattern often enough to name it: a French lender financing a Colombian project, a Colombian subsidiary paying royalties to a European parent, a management fee crossing the Atlantic — and the treaty rate never claimed, because the documentation was never assembled and the deadline passed quietly.

Treaty relief is not automatic. It requires residence certification and, increasingly, a defensible answer to the question of whether the recipient is the beneficial owner of the income or merely a conduit. That answer is far easier to give in advance than to reconstruct under audit.

What connects all three

These are three unrelated instruments — a minimum tax directive, a carbon border levy, a bilateral treaty. What they share is a direction of travel that has been consistent for a decade: European tax administration is moving from checking declarations to verifying substance.

The side-by-side safe harbour is granted on the basis of what a parent jurisdiction actually does, not what a structure declares. CBAM prices what a specific installation actually emits, not what an industry average suggests. Treaty relief attaches to who genuinely owns income, not to who is named on the payment instruction.

In each case the question is the same: can you evidence it? And in each case the evidence has to exist before the transaction, not after the assessment.

The failure we see most often

It is rarely aggressive planning. Aggressive planning tends to arrive with advisers attached and a file behind it.

What we see far more often is a structure that was correct when it was built and was never revisited. A holding company established when the treaty network looked different. A financing arrangement designed before interest limitation rules applied. A European entity with a registered address, a bank account, and no decision ever taken inside it — which was unremarkable in 2015 and is now the first thing an auditor tests.

None of this was wrong at the time. It simply stopped being right, and nobody scheduled the review that would have caught it.

Where this leaves a CFO

If your group is US-parented with Colombian operations and any European footprint, three questions are worth putting on the next agenda.

Does the side-by-side safe harbour apply to your group, and has each member state where you operate actually transposed it? The answer will differ by jurisdiction through 2026 and into 2027, and the gap between the OECD position and national law is where filing errors live.

If you export any covered commodity into the European Union, do you know whether you cross the fifty-tonne threshold, and can your production data withstand a customer's verification? That question is being asked now, by buyers, ahead of the 2027 surrender.

And on every recurring cross-border payment between Colombia and Europe — dividends, interest, royalties, service fees — is the treaty rate being claimed, and is the file that supports it complete?

None of these is exotic. All three are the kind of thing that is inexpensive to fix in advance and expensive to argue about afterwards. That asymmetry is the whole argument for reviewing them before the year closes.