This article is part of the GSMA’s Blueprint blog series, developed through direct engagement with its AI Taskforce of mobile network operators. The series captures real-world experiences from the telecommunications industry to help organisations navigate their AI transformation journey.
The Industry Challenge
The telecommunications industry continues to face a familiar challenge: turning Responsible AI (RAI) principles into everyday practice. Many organisations have commenced their RAI journey by doing things like publishing policy statements, setting up committees, and building governance structures … yet the gap between intention and execution often remains.  Frameworks are an important starting point, but their real impact comes when they’re paired with practical, on‑the‑ground implementation.
As AI systems scale, the real opportunity lies in embedding Responsible AI into daily operations and decision‑making, transforming it from a set of ideals into a foundation for resilient, trustworthy innovation. In practice, this means making Responsible AI part of the daily workflow, not just an added governance layer.
As part of our AI Transformation blog series, we’ll be taking a closer look at KT as a case study, since it illustrates the RAI maturity journey particularly well.
KT’s Structural Approach to Responsible AI
KT Corporation, one of South Korea’s largest telecommunications providers, has approached this challenge through structural reform rather than incremental policy adjustments. In 2024, the company created a position that signals how seriously it takes this issue: Chief Responsible AI Officer (CRAIO).
Establishing Operational Accountability
The CRAIO role at KT carries genuine operational authority. Rather than serving solely in an advisory or committee capacity, this role plays a central part in coordinating and guiding enterprise-wide AI initiatives. By pairing accountability with the ability to influence outcomes, KT reinforces its commitment to embedding responsibility into practice and not just policy. Empowering leaders with both responsibility and decision-making authority helps ensure ethical considerations are truly integrated into the development and deployment of AI systems.
The CRAIO role launched alongside KT’s Responsible AI Centre, operating within the company’s broader AICT strategy. The Centre runs two divisions that reflect the dual nature of responsible AI work. RAI Policy handles governance frameworks, ethical guidelines, and regulatory alignment – the structural foundation. RAI Tech manages technical implementation, risk assessment, and mitigation strategies – the operational reality. This separation acknowledges that responsible AI requires both policy coherence and technical capability working in tandem.
KT also established an external RAI Advisory Committee that brings in perspectives from outside telecommunications: academics, startups, and stakeholders. This structure recognises that internal teams can sometimes develop ‘blind spots’ in identifying potential harms or biases that may not be immediately obvious to those embedded in industry conventions. External voices provide a challenge and perspective function that internal teams often cannot always generate themselves.
Integrating Ethics into Development Processes
The CRAIO’s responsibilities span KT’s entire AI development lifecycle. The role involves developing ethical standards, but as importantly, ensuring these standards are applied consistently across the different departments and projects. This cross-functional coordination prevents responsible AI from becoming isolated within a single team or treated as separate from core operations.
KT included mandatory pre-deployment reviews for AI systems, a significant step introduced into the development phase. These reviews require leadership involvement at key decision points and distribute accountability across departments. Each system undergoes risk verification before deployment, positioning ethical considerations as fundamental requirements rather than final checkpoints.
This approach reflects a pragmatic philosophy: responsible AI considerations are significantly easier and less costly to address during development than after deployment. Problems identified in production require not just technical fixes but often reputational repair and regulatory response. By integrating these reviews into standard development processes, KT treats responsible AI as foundational to how technology gets built.
Contributing to Broader Standards
The CRAIO’s influence extends beyond KT’s internal operations. The role includes active participation in international AI regulation discussions, including Korea’s developing AI Basic Law and evolving regulations around personal data protection and copyright.
Through partnerships with organisations including Microsoft, GSMA, and others, KT contributes to shaping industry standards while learning from regulatory developments in other markets.
Linking to the GSMA Responsible AI Maturity Roadmap
The GSMA Responsible AI Maturity Roadmap helps organisations assess and develop their practices across four maturity levels – Foundational, Evolving, Performing, and Advanced – spanning five core dimensions: Vision, Operating Model, Tools, Third Party Ecosystem, and Change Management. Within this framework, the Executive Sponsorship sub-dimension examines whether leadership commitment extends beyond mere statements to drive actual organisational change.
KT’s experience shows what this looks like when it works. The company has moved beyond policy documents to create organisational structures, allocate dedicated resources, and establish processes that make responsible AI operational rather than aspirational.
For other telecommunications operators working to strengthen their responsible AI capabilities, KT’s approach illustrates how executive sponsorship translates into concrete organisational changes. The creation of the CRAIO role, the establishment of the Responsible AI Centre, and the implementation of mandatory review processes show how leadership commitment can fundamentally reshape how a company develops and deploys AI systems.
The Implementation Challenge
Creating a C-suite role specifically for responsible AI can establish clear accountability. KT’s approach offers one model for what transformation looks like when executive sponsorship moves from concept to implementation. It demonstrates that making responsible AI operational requires structural changes that redistribute authority and establish new decision-making checkpoints.
Conclusion – Looking Ahead
The telecommunications sector is far from unique in its rapid and expansive adoption of AI. The real issue isn’t whether organisations will encounter challenges with AI practices – they inevitably will. The key question is whether they’ve established the right structures and capabilities to recognise and respond to those challenges effectively. With the right foundations in place, organisations can navigate AI challenges more confidently while creating opportunities to drive innovation, build trust, and deliver long‑term value.