Fujitsu Limited today announced the introduction of a new enterprise-grade artificial intelligence platform that allows companies control over the development, deployment, and ongoing refinement of generative AI tailored to their internal operations.
The platform, set to begin phased rollout in Japan and Europe, promises a fully dedicated environment in which businesses can manage the full lifecycle of generative AI, from model creation and tuning to continuous learning and operational governance. Preliminary trial registration slated to begin in February 2026, ahead of a broader launch planned for July
Addressing Sovereignty and Security in Corporate AI
As enterprises increasingly deploy generative AI across functions such as research, customer engagement, and automation, concerns about data sovereignty and unintended information leakage have intensified. Fujitsu’s new platform is positioned as a response to these challenges.
Industry analysts say this launch comes at a time when corporate AI governance has moved to the forefront of enterprise technology planning. “Firms want both generative AI capability and assured security — particularly for high-value data,” noted one technology sector expert. “What Fujitsu is offering here is a way to balance those imperatives.” (This analysis is based on common industry commentary on enterprise AI needs.)
Integrated Features for Trust, Efficiency, and Scale
According to Fujitsu’s announcement, the platform integrates a suite of advanced capabilities designed to reduce operational hurdles for corporate adopters:
- Dedicated Secure Environment: Customers can host the platform in their own data centers or in Fujitsu’s private facilities.
- Embedded Trust Technologies: Proprietary scanners detect more than 7,700 vulnerability types and guardrail systems are deployed to prevent malicious exploits. Fujitsu also plans enhancements to counter the generation of inaccurate or misleading content (so-called “hallucinations”).
- High-Precision Models and Customization: The system centers on the Takane large language model, co-developed with AI partner Cohere, which supports strong Japanese language performance and image analysis. Built-in fine-tuning allows enterprises to refine models for their distinctive business contexts.
- Efficiency and Cost Reduction: Innovative model “lightweighting” and quantization techniques are said to cut memory usage by up to 94 percent, translating into more efficient computing and lower costs.
- AI Agent Development Tools: A low-code/no-code framework supports rapid construction of AI agents and promotes interoperability with existing corporate systems via Model Context Protocol and inter-agent communication capabilities.
Roadmap and Corporate Strategy
Fujitsu stated it will initially offer key platform features for trial from February to select enterprise customers, with staged releases thereafter leading to a full launch in mid-2026. The company also outlined plans to extend support beyond large corporate deployments to include edge and physical AI applications.
Also Read: 5 Reasons HDDs Will Continue to Dominate Enterprise Storage in the AI Era
