Digital sovereignty faces new challenges as AI may enable software reversibility, but legal and technical hurdles remain.
Digital sovereignty and software dependence
Digital sovereignty, a concept focused on maintaining control over digital resources, has increasingly centered on software reliance, with artificial intelligence possibly offering solutions for reversing such dependencies. This shift highlights a growing concern over how deeply government and public institutions depend on proprietary software solutions.Historical focus of digital sovereignty
For two decades, discussions on digital sovereignty have primarily centered on cloud computing, data, and infrastructure, often overlooking the deep reliance on software itself. Governments typically purchase software solutions with long-term contracts, leading to outdated systems, lost expertise, and reliance on a single supplier.Challenges of exiting software suppliers
Despite legal provisions for reversibility and data portability in public procurement, these are rarely implemented due to high costs. Exiting a software supplier involves complex tasks like system analysis, rule extraction, and migration, which are rarely feasible without significant resources.AI's potential in software reversibility
Artificial intelligence could make the process of analyzing and transforming complex software more manageable, reducing the cost of understanding and modifying applications. However, AI does not inherently make systems reversible; it may only make the concept more practical.Challenges of system migration
The transition of the National Education messaging system from Oracle's Convergence to a national solution highlights the difficulties of migration. Challenges include exporting address books, synchronizing calendars, retrieving email archives, and reconfiguring workstations, while preserving academic identifiers.Legal boundaries of software use
The European Union's Court of Justice ruled that software features, programming languages, and file formats are not inherently protected by copyright. This allows for independent implementation, but only within legal limits, preventing direct copying of code or circumventing licenses.AI and new dependencies
While AI could enhance software reversibility, it could also create new dependencies on foreign suppliers. Mass-generated code might be as opaque as existing software, emphasizing the need for understanding, documentation, and maintenance skills for true digital sovereignty.Strategic requirements for sovereignty
To ensure sovereignty, governments must prioritize the ability to rebuild or replace essential systems. This includes requirements like effective data portability, living documentation, regular system testing, and annual exit plans. Sovereignty is not just about owning software but retaining the ability to change it when needed.