Lately I’ve been hearing a lot about data sovereignty. Indian Tribes want data sovereignty. Countries around the world and state and local governments in the U.S. are talking about data sovereignty. Corporations are talking about data sovereignty, even though they use different terms to describe it.

The concept behind data sovereignty is simple. It means keeping data internally and not sharing it with the outside world, which today means keeping it away from the AI companies that are training their models on every bit of data they can put their hands on.

Corporate Data in the AI Cloud

Consider corporations. I’m positive that a huge amount of what happens inside corporations ends up in the AI cloud. It seems inevitable that employees at every company rely on AI to compose emails, analyze budgets, and write memos. Employees use AI notetakers to capture every word said in meetings. Even if a company has a policy of not doing any of this kind of work using public AI, you know employees are doing so privately to make their jobs easier.

Europe’s Push for Data Sovereignty

Entire countries are talking about data sovereignty, particularly in Europe. There is a quickly growing top-down movement on the continent to ban and replace large U.S. software tech vendors. We’ve recently seen the governments in France, Spain, Germany, the UK, and Switzerland begin the process to get rid of Palantir, a U.S. surveillance software vendor, and replace it with European software. Each country wants not only to protect data from U.S. tech companies but also from other countries. There is a lot of discussion in Europe about replacing many other U.S. tech firms.

Local Control in the U.S.

The data sovereignty effort in the U.S. is much more of a bottom-up movement. Many city government CIOs are demanding to have local control of citizens’ data. This partly comes from a rash of cities and local governments that have been hacked by ransomware attacks in the last few years. But this is also in response to citizens demanding that local governments protect their data.

Of course, there is opposition to the data sovereignty idea. For example, there is a tech coalition that has been pushing the Administration to create “freedom cities” that are exempt from federal regulations and where tech companies would have free rein. An effort to do this in Toronto failed a few years ago when citizens rebelled against the idea of creating a high-tech surveillance community.

Tribal Data Sovereignty

Indian Tribes might be ahead of everybody else on this issue, and there is a lot of discussion of keeping all tribal data sovereign from the rest of the world. Tribes have no interest in the federal government getting their local data and are even less interested in seeing tech companies profiting by selling the data to anybody willing to buy it.

A More Local AI Model

The idea of data sovereignty runs completely counter to the idea of building huge data centers that amass all of the knowledge of mankind. Sovereignty is really about privacy and logically leads to an AI environment comprised of local AI servers for corporations, local governments, and Tribes that are not connected to the cloud or to the big AI data centers.

Interestingly, the data sovereignty movement is not anti-AI. It instead comes from those who are against having their data gobbled up by a handful of big tech companies that want to monetize their data with no care for what that means to the people, companies, and governments that generate the most useful data. Most entities that shield themselves from big AI will likely use AI internally inside their own network. I think a whole lot of people have already seen the massive invasion of privacy that is an integral part of AI and are already looking for ways to opt out.