The telco of the future: Generative AI as a force multiplier in product strategy


  • Generative AI is driving significant disruption across the telco industry.
  • But to get the business benefit, telcos will have to look beyond low-hanging fruit use cases, and deal with more complex, capability-driven initiatives that deliver long-term value.
  • Firms that want to go a step further will embed this technology into product strategy; how much value-add they can achieve depends on their data and AI readiness.
  • For firms that don’t have data science talent, partnering is a good strategy, with partners bringing the AI capabilities and telcos the domain-specific know-how.
  • Key imperatives to transform from telco to tech-co include identifying high-impact use cases and breaking complex projects into short, sharp sprints that deliver exponential business benefit.

Generative AI is “a game-changer”, according to Deutsche Telekom CEO Timotheus Höttges. He and his executive team are revising product strategy to “exponentially develop” chatbots and call center support, among other initiatives that will redefine the firm.

With generative AI already creating value for firms across all industries, telcos must now revisit their product portfolios and move from generative AI point solutions to using it for complete business transformation. This will mean taking an ecosystem approach, building native AI products and solutions that combine the best of Internet of Things (IoT) and generative AI.

As low-hanging fruit AI initiatives are exhausted, investment in more complex generative AI projects that deliver long-term value will separate the winners and losers (Figure 1).

Figure 1. As the product portfolio gets more complex, so does the potential value generated

Figure 1. As the product portfolio gets more complex, so does the potential value generated

With the right approach to data and AI readiness, and investment in the partner ecosystem, telcos can make the move from traditional telco to “tech-co”, building more mature AI capabilities along the way.

From gen AI capabilities to product growth strategies

From our Generative AI Radar research, we’ve found that there are many use cases telcos are adopting as they make this shift, including:

  • Enhanced network maintenance.
  • Threat detection to strengthen cybersecurity.
  • Uplift in all stages of the software development lifecycle.
  • Set top box test automation.
  • Spectrum sensing through gen AI-generated synthetic data.
  • Network traffic analysis and anomaly detection.

However, this is the easy stuff. More exacting firms like SK Telecom are building out capabilities that extend across marketing, sales, process optimization, risk, and even sustainability. This level of innovation requires more data processing and business support. Some examples are:

  • Actionable insights for cross-sell/upsell opportunities (marketing).
  • Employee workflow assistants (personal productivity).
  • Risk modeling and analytics.
  • Predicting demand for products and services (insight generation).
  • Customer service chatbots.
  • Knowledge management, including document summarization and semantic search (content intelligence).

However, to get the value that AI promises, the telco of the future will have to integrate these capabilities into products and solutions, transforming the business by moving from low value-add to native AI-based products and services.

There are three product and service areas where generative AI can make a difference, at growing levels of maturity, business and technology complexity, and potential value capture:

  1. Enhance the existing connectivity portfolio (low value add): The natural choice for speed to market, organizational readiness, and cultural fit is to continue enhancing existing connectivity products using generative AI. Data collected from existing connectivity products could be used to increase the effectiveness and value of these products. For example, security offerings could use network data to predict and mitigate threats more accurately, or generative AI could be used for better network traffic analysis and anomaly detection (as mentioned above).
  2. Provide AI-ready infrastructure (medium value add): Another natural fit for telcos is AI-ready platforms or infrastructure. A platform approach could enable third-party app development in addition to native features, creating more of the personalized, feature-rich services of the future. For example, AI-enabled set top boxes could include capabilities such as object-based media, where individuals get an personalized version or program with branched narratives, and in sports, different camera angles or personalized advertising. Another platform, though not as obvious, is AI-ready data centers, where generative AI systems continuously optimize the operation of data centers by dynamically adjusting resource allocation, workload scheduling, and cooling strategies to maximize energy efficiency, minimize latency, and ensure high availability.
  3. Develop native AI-based services (high value add): As firms move from telco to tech-co, they develop and deliver AI-based solutions to solve deep business challenges. For example, telcos might have substantial data on fleets; using data collected from the vehicles, telcos can combine this with generative AI to monitor activity and optimize fleet utilization, providing a value-added service and new business growth opportunity. Another example is providing customer incident management services: generative AI can be used on raw log data such as alarms or program traces to help detect incidents; draft text for customer support requests or trouble reports (and converting them where necessary); and generate labeled clusters of categorized trouble reports for easy searching when similar tickets are raised, enabling faster and easier resolution.

Selecting the right services will depend on how ready certain telcos are for AI. Data is a critical factor: tech-cos, in this model, will have ready-prepared data for use in AI models.

This means making sure data assets are available, discoverable, accessible, and of high quality. At the strategic level, thought must be given to regional market needs and risk appetite, along with how well versed the C-suite is in the building blocks and challenges of a generative AI roadmap. Getting all this right means the telco can return higher value.

SK Telecom is already finding significant RoI from native AI solutions, including a solution called A-dot that provides call summary, real-time interpretation, and playlist creation and other entertainment for customers. By providing generative AI-tailored industry offerings (analysis of medical imagery in healthcare; personalized entertainment for media firms such as Netflix), along with horizontal solutions such as fleet management, security, and AI assistants, tech-cos will find new market potential, solve business pain-points, and reframe themselves as anytime-anywhere tech providers.

Figure 2 is an example of both vertical and horizontal solutions across the three stages of maturity already mentioned.

Figure 2. Fleet management is an example of a value-add, horizontal capability

Figure 2. Fleet management is an example of a value-add, horizontal capability

Build, buy, or partner?

With all this in mind, a big component of generating long-term value is the decision to build IP, buy IP, or partner. Again, the decision is based on general AI readiness.

  1. Build AI IP must consider both internal versus partner development:
    • Internal development is appropriate when there is resource and capability to design an AI-first operating model and the know-how to hire or develop talent to deliver native AI solutions.
    • On the other hand, bring in partners when know-how is limited and strategic partnerships are needed to deliver native AI solutions.
  2. Buy (or license) AI IP is necessary when firms are unsure of the long-term viability of integrated solutions, and firms such as OpenAI or Google provide strong governance capabilities. These capabilities can then be easily integrated into solutions.
  3. Partner for AI IP considers the wider partner ecosystem and value proposition strength. After considering go-to-market solutions, telcos can select partners that bring the AI capability and the telco brings expertise in IoT and domain-specific know-how.

Moving ahead

There are some key imperatives all telco firms should be aware of as they make the transition to tech-co, bearing in mind the specific build, buy, or partner approach taken.

  • The first is that although generative AI models are perishable, the data isn’t. This means that telcos should build abstraction into their architecture so they can choose which AI service provider they want, depending on performance.
  • Also, in working out which initiatives to onboard first, it’s important to articulate potential business value, and identify high-impact capabilities before implementation begins (Note: conducting a strategic value chain analysis of the business process can help address this challenge).
  • Another recommendation is that telcos create an AI foundry (to experiment and incubate new technologies) and an AI factory to foster extreme automation and turn learning from the AI foundry into new products, services, and applications. This will in turn balance and manage risks associated with AI evolution while scaling its adoption.
  • Finally, Rome wasn’t built in a day. Many firms we talk to start a complex-yet-high-value AI initiative, see lackluster returns, and go back to the status quo. To avoid this, think big, start small, and get started. A micro-change management approach is ideal, breaking big initiatives into bite-sized chunks for exponential business benefit.

We’re in the honeymoon period of generative AI. Telcos that grasp opportunity can gain market share. Now is the time for telcos to make this transformational new technology work for them, providing not only essential utility but significant value capture along the way.

Related Stories

Connect with the Infosys Knowledge Institute

Opt in for insights from Infosys Knowledge Institute Privacy Statement