In recent years, the focus in finance, public services, and healthcare has shifted from whether artificial intelligence (AI) can perform tasks like writing code to where that AI operates and what data it can access. IBM is responding to this shift by offering a self-hosted deployment option for IBM Bob, a platform that uses AI to assist with the development, maintenance, and modernization of software applications. This means companies can use the platform without having to send their code or data to an external service provider. Many businesses are adopting AI-based tools to streamline development, but in highly regulated industries, sharing sensitive code, customer information, or data from essential systems with external AI services is a significant concern. As a result, controlling where AI runs, what data it can access, and how it operates has become a key factor in its adoption. IBM has addressed this by enabling the deployment of AI capabilities directly within the environments where a company's data is stored. This allows development and modernization functions to occur without the need to transfer code or data to an external service. IBM Bob supports several deployment options, including on-site hosting, private clouds, sovereign clouds, and isolated environments that are not connected to the internet (air-gapped). Companies can use licensed AI models directly on-premise or in these isolated environments, or opt for a hybrid setup that connects to external AI services. Unlike many AI tools that are designed for cloud-based environments, IBM Bob can support the development and modernization of applications and data that are spread across multiple environments, from local systems to private clouds. The growing demand for data sovereignty — the control of data within national borders — is shaping the market. According to Futurum Research, businesses are expanding their use of hybrid and edge computing environments to maintain control over their data while still benefiting from AI technology. These hybrid and edge deployments are expected to make up 44 percent of the AI infrastructure market by 2030, compared to 46 percent for the public cloud. A June report by the IBM Institute for Business Value, titled The Calculus of AI Sovereignty, found that 68 percent of surveyed executives struggle to comply with national and regional data residency and sovereignty rules. Neel Sundaresan, general manager of AI and automation at IBM, explains the company's approach: "The future of AI in business will depend on security, governance, and sovereignty. Organizations need an AI that operates in environments they control, especially when they handle sensitive code and regulated data. The self-hosted deployment of IBM Bob allows companies to bring agentive AI directly into these environments, in order to benefit from this technology while maintaining security, compliance, and operational control."