TRC Publishes Guidance on the Use of Artificial Intelligence in Telecommunications Operations

In 2026 the Cambodian authorities issued guidance on the use of artificial intelligence in telecommunications operations. The guidance addresses the growing use of AI in areas such as network optimisation, fraud detection, customer service, personalisation, content moderation and cybersecurity, and articulates the expectations that apply to responsible deployment of AI in these operational contexts.

The guidance is relevant to all telecommunications operators active in Cambodia, and to vendors and service providers supplying AI-enabled solutions to those operators.

The guidance recognises the significant potential benefits of AI in telecommunications operations, including improved efficiency, faster detection of anomalies, better customer experience and more effective use of network resources. It also identifies risks that must be managed, including biased or unfair outcomes, opacity of decision-making, cybersecurity implications and unintended effects on subscribers or third parties.

The guidance covers AI applications used internally by operators, including tools that support decisions affecting individual subscribers such as service configuration, credit decisions, fraud investigation and personalised offers. It complements broader frameworks addressing AI at national level, including the national AI strategy and the cross-sector AI sandbox.

Operators are expected to have governance arrangements that provide senior-level oversight of AI initiatives, that assign clear responsibility for AI outcomes and that ensure appropriate involvement of technical, legal, compliance and customer-facing functions. Governance should extend across the lifecycle of AI systems, from initial design and data selection through deployment, monitoring and eventual retirement.

Accountability is a central concept. When AI is used to support decisions that affect subscribers, the operator remains responsible for those decisions and for their consistency with applicable rules. AI cannot be treated as a technical black box that operates independently of accountability.

AI systems reflect the data on which they are trained. Operators are expected to consider the quality, representativeness and appropriateness of training data, and to identify and mitigate potential sources of bias. Particular attention is warranted where AI is used in decisions that could disadvantage specific groups or that operate in sensitive contexts such as credit and fraud.

Data protection considerations apply to the collection, use and retention of data for training and operation of AI systems. Coordination with the applicable data protection framework is essential.

The guidance encourages transparency about the use of AI, including through information to subscribers where AI plays a significant role in decisions affecting them. Explanations of AI-supported decisions should be available to subscribers and, where necessary, to the regulator, in a form that supports understanding without necessarily disclosing all technical details.

Operators are also expected to be able to explain to themselves, and to their regulators, how their AI systems work at a level sufficient to identify issues and to make informed changes.

Human oversight is a central safeguard, particularly for decisions with significant effects on subscribers. Operators are expected to design their AI systems so that appropriate human review is available, so that individual decisions can be reviewed by qualified personnel when required and so that patterns of concern can be identified and addressed.

The intensity of human oversight should be calibrated to the significance of the decisions supported by the AI system. Highly automated processing of routine matters may require limited oversight, while decisions affecting service continuity, credit or fraud investigation require more substantial human involvement.

AI systems evolve over time as data changes and as the environment changes. Operators are expected to monitor the performance of their AI systems continuously, to identify drift and unexpected outcomes and to update systems as required. Post-implementation review supports learning and continuous improvement.

Aggregate monitoring of the impact of AI systems on subscribers, on network performance and on operations helps operators identify systemic issues and opportunities for improvement.

Operators frequently use AI capabilities supplied by third parties, including equipment vendors, cloud providers and specialist AI companies. The guidance addresses the responsibilities of operators in these arrangements, including due diligence on suppliers, contractual protections for data and AI outcomes, monitoring of supplier performance and contingency planning for the loss of critical AI capabilities.

Coordination between operators and suppliers on the design, deployment and support of AI systems supports responsible outcomes and helps manage risks.

AI systems are themselves targets for cybersecurity threats, including attempts to manipulate their behaviour, to extract sensitive data or to disrupt their operation. Operators are expected to include AI systems in their overall cybersecurity arrangements and to consider the specific risks associated with AI, such as model theft, adversarial inputs and data poisoning.

The interaction between AI, operational technology and network resilience should be considered in cybersecurity planning and in incident response arrangements.

For operators, the guidance calls for a review of existing AI initiatives, of governance arrangements and of vendor relationships, to ensure alignment with the expectations. Cross-functional coordination, investment in AI literacy across the organisation and engagement with the regulator on significant deployments support responsible use of AI.

For vendors and specialist providers, the guidance sets expectations that should be reflected in the products, services and contractual arrangements offered to Cambodian operators. Providers that engage constructively with the guidance and support their customers in meeting the expectations are well positioned to build long-term relationships.

Lex Civora advises telecommunications operators, vendors and enterprise customers on the interpretation and application of the AI in telecom operations guidance, on the design of governance and vendor arrangements and on the alignment of AI initiatives with the broader legal and regulatory framework applicable in Cambodia.