The objective of this contribution is to analyse the legitimacy of the OECD’s work on Pillar Two. The starting point is that the effectiveness of the new global minimum tax is clearly based on its so-called ‘devilish logic’. As such the project relies heavily on expert knowledge that is supposed to guarantee output legitimacy. At the same time, the consensus reached in the Inclusive Framework (IF) is supposed to bless the global minimum tax with a form of input legitimacy. Nevertheless, the contribution comes to the conclusion that the legitimacy of the OECD’s work on Pillar Two is falling short. The central point is that the governance process of the OECD should meet burdensome standards of ‘good’ governance including accountability (i.e., throughput legitimacy). Unfortunately, the political accountability and the technical accountability of the work on Pillar Two were clearly insufficient. The OECD, and the Centre for Tax Policy and Administration (CTPA) in particular, was not appropriately operating ‘in the shadow of politics’ and the ingenious solutions of the global minimum tax were not sufficiently scrutinized by independent technical experts. As a result of these shortcomings, the legitimacy of the work on Pillar Two eventually relied too strongly on expert knowledge (the ‘devilish logic’) and therefore on output legitimacy only. This conclusion illustrates once more the urgent need to rethink the legitimacy of international tax governance in general and the role of the OECD in particular.