Agentic AI: The Next Big Thing in Telecommunications?

November 2024

Telecoms & Connectivity

Molly Gatford

Senior Research Analyst

Agentic AI is an emerging concept in the field of artificial intelligence that has recently gained significant prominence. The technology promises to offer new capabilities, such as the ability to adapt, learn, and make independent decisions with minimal human oversight required.

Agentic AI systems typically leverage multiple LLMs (large language models) that are deployed across various applications to enable autonomous decision making and action-taking and the completion of tasks with complex workflows.

Whilst generative AI has been advertised as a tool to automate various functions within the telecommunications industry, this ability is limited by the data with which the AI system is provided. However, Agentic AI will fulfil an unmet need in the telecommunications sector, by automating processes that enhance operational efficiency across the industry.  

With the advancements that agentic AI can provide over both traditional AI and generative AI, Juniper Research anticipates that we will see it gain increased attention from players in the telecommunications industry over the next year. Autonomous decision making will enable better network management, customer service over chatbots, and more responsive security over networks.

Agentic AI vs Other AI Systems

Source: Juniper Research

The autonomy of Agentic AI will ensure that it offers a clear benefit over other AI technologies. But what is the specific relevance of the technology for the telecommunications industry?

Benefits of Agentic AI in Telecommunications

There are several different ways in which agentic AI systems could be utilised in telecommunications; offering advanced solutions for both network operators and enterprises.

For network operators, agentic AI will be used for the following:

• Network Management and Optimisation: Agentic AI could be used to optimise the performance of networks and reduce latency, by monitoring network traffic and automatically adjusting parameters. The predictive capabilities of agentic AI will also allow it to predict when a network failure will occur; in turn minimising downtime by ensuring proactive maintenance.

Agentic AI will aid in the monetisation of 5G networks for real-time IoT applications. Agentic AI will not only help with the management and optimising of the growing number of applications connected to 5G networks, but could help to ensure that applications receive a consistent performance from the network.

• Network Security and Fraud Detection: Agentic AI’s ability to adapt based on its learning will allow it to provide enhanced network security solutions within the telecommunications industry. Where current AI systems can monitor patterns in network traffic and identify when a change in these patterns occurs, agentic AI can go one step further and adjust security protocols based on the threat that it identifies.

For enterprises, agentic AI could be used for the following:

• Virtual Assistants: Agentic AI will enable virtual assistants to autonomously perform multi-step tasks; taking actions across multiple applications or services based on a user’s request. These virtual assistants can be used for both customer care and conversational commerce use cases, where they can make autonomous decisions based on customer requests and take action accordingly.

For conversational commerce interactions, these virtual assistants will be able to handle shopping tasks, including the process of searching for products and completing the purchase on behalf of the customer, with appropriate authorisation; creating a more secure customer experience.

• Personalisation of Customer Experiences: Agentic AI also has the potential to predict when a customer will most likely respond to a messaging campaign. The technology will be able to predict changing user preferences, such as a preferred channel for communication, based on patterns and can adapt its approach without the need for human intervention.

By employing agentic AI within business messaging applications, businesses will see increased engagement with messaging campaigns, as the technology will more accurately predict the channel with the highest engagement and the time to send a message to each customer.

Juniper Research therefore expects vendors that provide operator and enterprise solutions to be first to invest in agentic AI for use cases in telecommunications. Those that have already invested in generative AI must look to offer agentic AI as a complementary technology for enterprises, to provide solutions which further improve operational efficiency and boost customer experience.  

We believe that agentic AI will be effective when integrated as part of operator networks to ensure efficiency and sustainability of networks. Improving network efficiency will be key to operators maximising revenue from 5G and future 6G networks.  

Telcos are already exploring and implementing AI technologies, however, there is not yet widespread adoption.  Juniper Research anticipates that the first agentic AI solutions for telcos will become commercially available in 2026. To effectively improve network management, agentic AI will have to be deployed across several key systems. This will include integration within the core network systems, the operations and management systems, and the data and security systems.