Tax Notes State · 3 August 2026
Taxing the Black Box: algorithmic opacity and a California AI tax doctrine
Published in Tax Notes State: when opaque AI systems materially shape taxable activity, who should bear the resulting informational risk?

My article “Taxing the Black Box: Algorithmic Opacity and a California AI Tax Doctrine” was published in the August 3 issue of Tax Notes State.
The question
When opaque AI systems materially shape taxable activity, who should bear the resulting informational risk: the taxpayer or the tax authority?
The argument
California does not need a new “robot tax” to answer this question. Its existing legal framework can support:
- an algorithmic nexus standard;
- an opacity-adjusted allocation of evidentiary burdens for AI-intensive taxpayers;
- reciprocal transparency duties when the state itself uses AI in tax administration.
Why it matters
For businesses that rely heavily on AI, the allocation of the burden of proof can be as decisive as the rate of tax. The article suggests how that allocation can remain balanced on both sides.
My thanks to Andrea Muse and the Tax Notes State editorial team.
Published in Tax Notes State, August 3 issue.
