AI for connectivity: how policy makers can help digitalisation

Artificial intelligence (AI) systems have significant potential to improve the operations of communications networks, including those offered by communications service providers (CSPs) and also by end-user organisations (including enterprise and governments). Although the adoption of AI in networks has begun, the communications industry has significant progress to make in order to fully realise the potential of AI. Policy makers, including governments and regulators, have a role to play in facilitating this adoption, and maximising the benefits of AI in communications networks. 

Analysys Mason was commissioned by Cisco to set out recommendations for how policy makers can help network operators to overcome challenges to embrace the capabilities of AI, which will drive a range of benefits in support of countries’ wider digitalisation goals.

The paper first explains the role of digital communications networks in supporting wider digitalisation goals. Most national and regional governments have identified the benefits of digitalisation to the societies and economies that they govern. The goals of digitalisation include economic growth, increasing productivity and skills, improved delivery of public services, production of intellectual property, improving competitiveness with other economies, social inclusion and sustainable living. Any improvements in the operation and performance of digital communications networks will further support these goals. 

The paper then focuses on the network operation benefits provided by a selection of the most promising network-facing AI use cases, including:

• anomaly detection, root cause analysis and issue prioritisation
• configuration fidelity
• predictive maintenance
• capacity management and planning
• capex optimisation
• optimisation of wireless networks
• energy consumption optimisation.

The use cases deliver a rich combination of benefits across four important areas: improved resilience, improved efficiency, improved performance and savings in energy consumption. By improving the operations of networks, AI helps to realise the benefits of digitalisation.

However, despite the potential benefits of AI applications in communications networks, implementation is being held back by a range of barriers. Data is often held in silos, which takes significant resources to prepare before it can be used by AI. Operators have concerns trusting AI to influence the configuration and security of their networks. Network providers are typically staffed by network experts, and may lack the critical data science and coding skills to implement AI. Demonstrating a return on investment for AI business cases can be complex, which can hinder securing investment budgets. Additionally, regulatory compliance burdens and uncertainties about meeting requirements may create barriers.

With these barriers in mind, the paper proposes a strategic framework of recommended policy initiatives under the following four high-level themes:

• Engage: policy makers should first establish a dedicated market monitoring function. This function diligently tracks AI developments in communications both locally and worldwide, gathering intelligence through research, professional advisers, interviews and conferences. Recognising the power of collaboration, policy makers should then convene industry stakeholders to share knowledge and best practices. These gatherings include governments, regulators, network providers, technology vendors, end-user representatives and academics.
• Facilitate: policy makers can help to develop ‘soft’ instruments such as frameworks and  standards. These tools should focus on key themes like security, safety, risk management, data management, transparency and accountability. Policy makers should map current skill shortages, and address these through government-supported training programmes and financial incentives. Looking ahead, they should ensure university courses integrate AI and communications modules.
• Implement: policy makers should define clear research and development (R&D) programmes, including for new AI infrastructure, fostering partnerships, or researching specific applications. Transparent project criteria and published budgets attract innovative applications. Governments can incorporate AI into their own networks, building on existing policies to foster the use of AI in general, and also on the learnings developed by the actions above.
• Intervene: policy makers can offer incentives like tax relief and government-supported finance facilities. These incentives should be tied to specific policy aims, such as using technology from R&D programmes or deploying certain AI features. Finally, policy makers should ensure that any AI-related regulations are outcomes based and forward looking. 

By implementing these recommendations, policy makers can effectively foster the integration of AI in communications networks, driving innovation and ensuring a secure, efficient and sustainable digital future.