The fundamental tension between GILTI and GloBE is structural. GILTI (or its successor for groups subject to Pillar Two) is a blended CFC regime: it pools income and taxes across all controlled foreign corporations and applies a single blended rate. GloBE works jurisdiction by jurisdiction: an ETR below 15% in one jurisdiction triggers a top-up tax for that jurisdiction, regardless of how well-taxed other jurisdictions in the blended pool are. A group that comfortably passes a blended GILTI test can still face GloBE exposure in individual low-tax jurisdictions.
How the blending problem arises
Consider a US-parented group with CFCs in three jurisdictions:
| Jurisdiction | GloBE income (€m) | Local tax (€m) | Jurisdictional ETR |
|---|---|---|---|
| Ireland | 100 | 12.5 | 12.5% |
| Germany | 80 | 24.0 | 30.0% |
| Singapore | 60 | 10.2 | 17.0% |
Under a blended approach, the combined ETR is roughly 19.4%, well above 15%. Under GloBE, Ireland triggers a top-up tax of €2.5m (the shortfall from 15%). GILTI, applying a blended rate, would not produce a charge on the Ireland income at the same rate that GloBE does, as it absorbs the Irish shortfall against the German over-tax.
The Side-by-Side Safe Harbour: what it does
The Side-by-Side Safe Harbour, introduced in September 2026 Administrative Guidance, is designed to treat GILTI (or a qualified successor blended CFC regime) as satisfying GloBE obligations for entities within the blended pool, subject to conditions. Where the safe harbour applies, the GloBE jurisdictional ETR test is not run separately for the entities in the blended pool. The blended regime is accepted as the proxy compliance mechanism.
The safe harbour does not eliminate the GloBE framework. It creates a deemed compliance outcome for entities within scope. Entities outside the blended pool (domestic US entities, joint ventures, non-CFC structures) are not covered and must be assessed under standard GloBE rules.
Conditions for the safe harbour
The safe harbour applies where all of the following are satisfied:
- The UPE is located in a jurisdiction that operates a qualified blended CFC regime (currently the United States with GILTI, or a jurisdiction with an equivalent mechanism formally approved by the Inclusive Framework)
- The blended CFC regime applies to substantially all the CFCs of the group for the period
- The effective rate of the blended CFC regime, computed on a GloBE-consistent basis, is at least 15% (the adjusted effective rate test in the Administrative Guidance)
- The group files the required disclosure in the GIR and in the blended CFC regime return
The adjusted effective rate test is the most technically demanding condition. It requires recomputing the GILTI effective rate using GloBE income figures rather than US tax figures, stripping out certain US-specific adjustments (the GILTI inclusion percentage, the foreign tax credit haircut) and substituting GloBE-compliant income and tax amounts.
Where the safe harbour does not resolve the blending tension
Even with the Side-by-Side Safe Harbour available, three scenarios remain where GloBE exposure is not eliminated by the blended CFC framework:
- The adjusted effective rate test fails. If the GILTI pool, recalculated on GloBE terms, falls below 15%, the safe harbour does not apply and jurisdiction-by-jurisdiction GloBE analysis is required in full.
- Entities outside the blended pool. US domestic entities, minority-held entities below the CFC threshold, and entities in jurisdictions excluded from the blended pool are assessed under standard GloBE rules.
- QDMTT jurisdictions. Where a jurisdiction has enacted a QDMTT, the QDMTT applies regardless of the safe harbour. The QDMTT collects the domestic top-up tax locally, with the safe harbour reducing (but not necessarily eliminating) any residual UTPR exposure.
Planning implications for US-parented groups
The safe harbour gives US-parented groups a viable path to GloBE compliance without full jurisdiction-by-jurisdiction computation for entities in the GILTI pool, but it is not automatic. Groups need to:
- Run the adjusted effective rate test before electing into the safe harbour. A group that elects in and subsequently finds the test fails faces a re-filing obligation.
- Identify entities outside the GILTI pool. These are the residual GloBE risk. Map them and model their jurisdictional ETRs independently.
- Coordinate the GIR disclosure with the GILTI return. The safe harbour requires consistent disclosure in both filings. Mismatches between the GIR and the US tax return will attract scrutiny during peer review and information exchange.
- Monitor Inclusive Framework approval of GILTI successor regimes. US tax reform may alter the GILTI structure. The safe harbour's availability depends on the successor regime meeting the approval criteria.
Pillar2OS handles the blending/jurisdiction tension directly in the computation layer: the adjusted effective rate test for the Side-by-Side Safe Harbour runs as a parallel computation alongside the standard GloBE ETR, flagging entities inside and outside the GILTI pool and producing the GIR disclosure fields automatically. See the Computation Centre.
Related
This article is a companion to the earlier deep-dive on the Side-by-Side Safe Harbour for US-parented groups and the UTPR in 2026 update. Groups assessing GILTI blending should also read explicitly conditional taxes for the covered tax classification implications.