Who controls the data, controls the future
AI now sits inside finance, defense, health care and public administration, and it runs on two things most governments do not actually control: the data that train it and the compute that processes it.
In the modern digital era, data and compute power have become the strategic assets of nations, akin to ports or power grids. The control over these assets is crucial, as the future of nations may be decided in boardrooms where they hold no seat. The shift in AI's role from a mere commercial product to an integral part of sectors like finance, defense, healthcare, and public administration underscores this reality.
However, the entities that govern these assets, the data and the compute processing them, are often beyond governments' control.
AI's influence now extends into critical sectors, raising concerns about sovereignty. Sovereignty, in this context, can be categorized into three aspects: data residency, compute sovereignty, and governance sovereignty. Currently, most governments focus on data residency, the practice of keeping citizens' data within national borders.
This aspect is the simplest to legislate and is the primary focus of most policy. However, compute sovereignty and governance sovereignty are more challenging. Compute sovereignty deals with who owns the chips and data centers for processing, while governance sovereignty concerns setting the rules for AI systems, rather than merely adopting the standards set by dominant developers.
While many countries have recognized the importance of AI sovereignty and have taken steps towards it, their efforts often create a fragmented regulatory landscape. This situation, although seemingly sensible as it diversifies away from major US and Chinese firms, risks creating a patchwork of national fortresses that cannot communicate with each other.
The lack of coordination between these sovereign AI efforts could lead to a race to the bottom, where jurisdictions compete for the most permissive AI rules instead of the safest ones. This scenario could potentially undermine the security and effectiveness of AI systems worldwide.
Written by urgent.news from The Jakarta Post's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
