Governments push toward sovereign AI despite data and security hurdles

National governments are moving from AI experiments toward controlled deployments that keep models, data and outputs within sovereign boundaries. IDC’s Alan Webber says the biggest barriers are data readiness, governance, skills and cybersecurity.

Governments are beginning to move AI from pilots into real deployments, though progress remains uneven across civilian, defense and intelligence organizations. Alan Webber, program vice president for national security, defense and intelligence at IDC, described adoption as still early, with agencies working through data quality, model tuning, policy maturity and cybersecurity concerns.

Sovereign AI refers to systems where the model, data and outputs are created, controlled and used within a defined political or geographic boundary. Unlike global models such as Claude or ChatGPT, sovereign systems are designed to keep sensitive government data from flowing across borders and to reduce reliance on externally developed models. Webber said the approach is most visible in national security, defense, intelligence, social services, health care and financial services for citizens.

The largest obstacle is data, including whether agencies know what they hold, whether it is accurate and how it should be used. Governments also need rules for access, auditing and appropriate use, along with new workforce skills for building repeatable AI capabilities. Cybersecurity remains another unresolved challenge, with Webber warning that current protections are still being built as the technology advances quickly.

52

Impact Score

Flexible data centers could ease grid bottlenecks

Startups, utilities and chipmakers are testing ways for computing facilities to reduce electricity use during grid stress. The approach could speed connections, but critics warn it cannot replace new generation and transmission.

AMD and Rackspace plan dedicated AI compute rollout

AMD and Rackspace have finalized a phased deployment for dedicated AMD-based compute across Rackspace data centers. The capacity is aimed at regulated enterprise workloads, including clinical AI and large-scale inference.

Lexar tests SSD offloading for local AI models

Lexar is developing an AI-focused SSD approach designed to cut DRAM demand when running large language models on consumer PCs. Internal tests show the company’s storage offloading can load models that traditional local frameworks struggle to run with limited memory.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.