In an interaction with Gulveen Aulakh, global chief technology officer (CTO) Erik Ekkuden said some fraud prevention solutions being built in India could be taken to global markets. The same characteristics would take India ahead of the world, in adopting and democratising Agentic artificial intelligence (AI). The top executive also advocated for a shift in mindset so that net neutrality and technology evolution enable each other. Edited excerpts:
The Prime Minister has called for a global framework for fraud prevention. What signal does that send, and is it an opportunity for Ericsson?
It is an opportunity. Fraud prevention and security for transactions are among the first network capabilities that can be scaled. Network application programming interfaces (APIs) can support fraud prevention for banks and financial services, and the next step can include advanced positioning and quality-on-demand services. This aligns with the central government’s ambition for secure services.
It can combine systems such as Aadhaar, Unified Payments Interface (UPI) and digital public infrastructure with real-time information from the network.
To identify a fraudulent call or transaction, the assessment needs to happen in real time. Ericsson can provide real-time network data and connect the network to applications through APIs. Static backend systems remain important for identity and other records, but dynamic decisions require live network information.
We are aligned with India’s direction and can work with telecom service providers (TSPs) in India while helping develop services that other markets also need.
Fraud prevention is a major challenge for financial institutions worldwide, including in Brazil, the US, Europe and Australia.
Could India be an ideal market to develop and scale these solutions?
India should be among the leading markets pioneering these solutions. It is in good company with other advanced markets. Working together, these countries can move forward on services such as quality on demand, where applications receive different levels of network performance depending on their requirements. India and other leading markets can help solve these challenges.
Do you see India among the early adopters of Agentic AI?
There is a lot of innovation in Indian universities and the startup ecosystem, although it is too early to say how much of it is scalable. The knowledge and ambition are there. India’s work on AI, including what we saw at the AI Impact Summit, suggests that the country can lead in AI applications and in distributing AI more widely across society. This is often described as democratising AI. AI cannot run only in large data centres; it also needs to move into homes and devices and be trusted. Ericsson can help India and work with local startups to bring those applications onto the intelligent fabric, with the required security, trustworthiness and performance.
India saw the world’s fastest 5G rollout. Will it help India prepare for AI-native 6G?
Yes. Some global hubs are ahead in areas such as humanoids and agentic AI, but India has an opportunity to combine strong 5G networks with its AI ecosystem. That can help Agentic AI and physical AI flourish. The transition towards AI-native 6G will be an evolution, not a reason to wait.
Does network slicing mean India needs to rethink net neutrality?
The important point is that mobile networks have continuously improved the experience for everyone, from 2G through 5G. That general Internet access must continue, and everyone who has Internet access should retain it. On top of that, networks need the ability to tailor performance to the requirements of different services.
An uplink-heavy wearable, for example, may have very different needs from another consumer or enterprise service. Treating all services as if they were the same could deprive users of a reasonable experience. We need to accommodate both what technology can do and what applications need.
That may require a shift in mindset, but it is a natural evolution as technology and user demand develop.
What are Ericsson’s manufacturing goals in India?
Ericsson has been a part of India for a long time and has substantial research and development (R&D) in Bengaluru, Chennai and the Delhi-NCR region.
We work with leading universities, including IITs and IISc. This has helped us expand development, including AI hardware in Bengaluru. It has also enabled us to manufacture in India not only for India but for exports.
We build 5G products for India in India and export components to Europe, including 9 million components this year, as well as 15,000 5G antennas exported from India to Europe. This illustrates why India matters not only to Ericsson but also to the wider world: The country can build capabilities for global markets.
How far can manufacturing in India expand across the value chain, especially for components?
That will not happen by itself; it takes time and it is an area that the government recognises. The ecosystem needs to develop. We are trying to bring suppliers to India so that they can establish operations and grow their businesses for both India and global markets.
Moving established capabilities from elsewhere takes time, but in the long term, there is a significant opportunity for more advanced components to be designed and manufactured in India.
India has recently come online with Aduna, the global network API platform, which now includes three Indian telcos. Can India be a proving ground for applications that can then be taken to other markets?
Yes. If an application works in India at the right price point, scale and service level, it is likely to work in many other markets.
India has a scale that makes other markets look small. There can be local adaptations, particularly in regulated sectors, and network maturity differs among countries. Indian networks are good and at the forefront in areas such as 5G Standalone, network slicing, programmability and AI-powered 5G. Not every network globally is as advanced, although others will get there. Some markets, including Singapore, are more advanced in particular areas, and that creates opportunities for learning in both directions.
Has the Indian telecom industry taken too long to adopt 5G network slicing?
Other countries have used slicing for mission-critical services, such as police body cameras, and in consumer settings, including stadiums and sports events in Singapore. The technology has matured over several years, and early leaders are already using it commercially for consumers and enterprises. Indian networks are ready, devices are available, and launches are now taking place. It is a good time to accelerate differentiated connectivity because there is demand from consumers, enterprises and government services. The benefit of network slicing is that one 5G network can support multiple virtual networks or service categories without requiring additional spectrum or base stations.
Beyond connectivity, what other capabilities could a 6G slicing-enabled network support?
AI-native 6G will need to provide tailored services and performance at scale, not only for individual users and applications but also for agentic workflows. They could support advanced and indoor positioning, sensing, security services for banks and financial institutions, fraud prevention and AI inference at the edge.
AI assistants could help people avoid fraud or translate languages in real time. Network capabilities could potentially help handle India’s many languages automatically.
There could also be visual recognition for objects at home. Networks will do more than simply pass bits, including providing guaranteed uplink performance.
