They say a week is a long time in politics; based on recent events surrounding AI, the same can be said of technology.
In the short time since our last article, 2025 Future Gazing, the cost of Stargate, a US initiative to build AI supercomputer data centers, ballooned from $100bn to $500bn. Then, a few days later, DeepSeek, a Chinese AI start-up, launched an open, cheaper, more efficient, but equally effective chatbot, R1, on simpler infrastructure at a fraction of the cost of ChatGPT and others.
This brought into question the accepted AI cost model. On January 27, the value of Nvidia, a leading supplier of AI infrastructure platforms and the darling of US stock markets since the release of ChatGPT in November 2022, sank by over half a trillion US dollars—the largest single-day loss in stock market history.
Since then, David Sacks, Trump’s AI and crypto “czar,” and OpenAI have confirmed there’s evidence that DeepSeek used OpenAI’s AI models to train R1, a process that violates OpenAI’s terms of service and equates to theft (not that OpenAI is squeaky clean). Watch this space! The geopolitical stakes are high as the United States and China tussle for AI supremacy.
But what does this mean for tax?
Our last article described how AI has the potential to be the most impactful technology ever, particularly for tax due to its knowledge and information-based nature. On January 25, The Economist said, “At the very least, AI will add to the next decade’s productivity gains, fueling economic growth. At the most, it will power humanity through a transformation comparable to the Industrial Revolution.”
Be that as it may, the Industrial Revolution was built on physical (tangible) assets like coal, iron (then steel), railways, machinery, canals, roads, textiles, steamships, and communications networks. By contrast, AI is upending the world of intangible assets like scientific knowledge, intellectual property, financial systems and banking, skills and labor training, legal and political frameworks, education and literacy, and organizational management. As such, AI operates squarely in our space—so, whatever impact AI has in general, expect that to be heightened for tax!
_______
Change on this scale is bound to raise concerns (our last article invoked a few fearful reactions), but that may be unwarranted. It all depends on how we respond to AI, and there are reasons to be (very) optimistic.
On January 24, the Tax Law Technology Center at the WU (Vienna University of Economics and Business) posted on LinkedIn describing how AI-driven innovations are already reshaping how tax authorities and businesses handle compliance in five ways:
- Enhanced data analytics: the regulation-related analysis of massive data sets
- AI-powered auditing: efficient analysis of returns and financial records
- Streamlined record-keeping: better documentation powered by blockchain and cloud computing
- Real-time reporting: automated, dynamic, (near) real-time reporting systems
- Global collaboration: common international reporting standards
These are important, but as tax relies ever more on data held on digital platforms, more fundamental issues have emerged. For historical reasons, unreliable or “bad” data are pandemic in our industry, and our effectiveness as a profession faces erosion due to the continuous struggle with subpar source data and the enterprise technology that supports it. So far, the following has largely eluded modern tax teams:
- A comprehensive tax data model and data map: the ability to understand and measure how the data in our systems represents and reflects the myriad tax-relevant events that occur across our organizations
- A complete tax solutions guide: the ability to holistically manage and control the systems that support and ensure the ongoing quality of those data models
- Meaningful skills succession: the few uber-talented individuals who master the above (sometimes called “unicorns”) have yet to learn how to pass on that knowledge or establish it as a transferable profession.
Consequently, tax is on the back foot, trapped by its “day job” of surviving the basics of compliance within increasingly tech-driven ecosystems. Unicorns offer only temporary respite.
This undermines the ability to engage in higher-value activities and hampers the adoption of nascent tools to build new capabilities based on extensive system visibility, enhanced operations transparency, and exhaustive insights from properly governed and trustworthy data. Tax belongs in this world, and AI offers a path that is unforeseeable by other means.
However, don’t get excited yet. The task remains daunting, and AI may need to evolve a generation or two before this is realistic. That should be available next week then!
Until next time …
No comment yet, add your voice below!