What’s up with… Ericsson, Singtel, AI threats

• Ericsson offloads iconectiv for $1bn
• Singtel teams with Bridge Alliance on GPU-as-a-Service
• AI presents more than 700 risks, according to the MIT

In today’s industry news roundup: Ericsson will bank $1bn from the sale of iconectiv, its number portability unit, to Koch Equity Development; the Bridge Alliance is helping to take Singtel’s GPU-as-a-Service offering to the broader Southeast Asia region; the MIT has identified more than 700 threats that come with the use of AI;  and much more!

Ericsson is to bank 10.6bn Swedish Krona ($1bn) from the sale of its Bridgewater, New Jersey-based number portability and digital identity unit iconectiv to investment firm Koch Equity Development (KED). The Swedish vendor currently holds a majority stake in iconectiv, which it added to its empire as part of the 2012 acquisition of Telcordia, while Francisco Partners owns a 16.7% stake. The transaction is expected to be completed during the first half of 2025. “iconectiv serves over 5,000 customers across various sectors as a leading provider in number portability solutions, and a provider of core network and operations management, numbering and data exchange services as its main segments, and with limited strategic synergies with the rest of Ericsson’s portfolio,” noted the vendor in its announcement about the deal. According to KED, “iconectiv is the authoritative partner to many of the largest communications service providers in the US and international markets. Its core network and operations management, numbering and data exchange services span trusted communications, digital identity management and fraud prevention. The world leader in providing number portability solutions, iconectiv is the number portability administrator in 10 countries, including the US,” and works with the Alliance for Telecommunications Industry Solutions (ATIS) to serve as “the Secure Telephone Identity Policy Administrator (STI-PA) for the US, a critical role in industry efforts to mitigate illegal robocalling.” Richard Jacowleff, iconectiv’s CEO, noted that the company "is excited to join KED along with the resources and capabilities they have to support our next chapter of growth. This acquisition marks a significant milestone in our journey, aligning us with a shareholder focused on long-term value creation that extends our track record as a trusted partner to service providers and regulators in the communications sector." KED noted that the acquisition of iconectiv underscores its “continued investment focus on telecommunications software and data services, building on the successful 2021 acquisition of Transaction Network Services (TNS). With KED's resources and capital, iconectiv will be well-positioned to unlock both organic and inorganic growth opportunities,” it added. 

Singtel has teamed up with the Bridge Alliance, which has 35 member operators worldwide, to bring Singtel’s GPU-as-a-Service (GPUaaS) offerings to enterprises across Southeast Asia. Some of the region’s major telcos – namely AIS, Maxis and Telkomsel – “are among the alliance’s early adopters to cater to the growing demand for AI computing in Thailand, Malaysia and Indonesia respectively, having signalled their interest in the service,” noted Singtel, which teamed up early this year with AI chip giant Nvidia to develop AI computing services and announced in March that it would launch GPUaaS services during the third quarter of this year. As part of the agreement, Bridge Alliance member operators will get access to the GPUaaS offerings from Singtel. “As the GPU needs expand in these countries, more GPU clusters will be launched into the respective countries to accelerate and scale the businesses,” noted Singtel. The Bridge Alliance will help its members with their “go-to-market strategies and effectively drive the success of the GPUaaS offerings with the enterprises,” it added. Bill Chang, CEO of Singtel’s Digital InfraCo unit, stated: “We are very heartened to see such strong interest from AIS, Maxis and Telkomsel to partner us in our GPUaaS offerings. Our collaboration with Bridge Alliance and telcos in the region will help democratise and accelerate the use of AI by enterprises across all industries, giving them the tools to achieve greater productivity and business value with our next-generation digital infrastructure and solutions. This reinforces our goal of being a catalyst for innovation and supporting the digital transformation of Southeast Asia’s digital economies.” Read more. 

The debate about whether AI is a boon, or a bane, to mankind continues following the publication by the FutureTech group of the Computer Science and Artificial Intelligence Laboratory (CSAIL) at the Massachusetts Institute of Technology (MIT) of a new online public database, the AI Risk Repository. Designed as a resource for organisations considering the deployment of AI, it lists and explains more than 700 potential hazards and outright dangers they could face when scrambling aboard the AI bandwagon (fools rush in, and all that..). MIT claims the AI Risk Repository is the “most comprehensive source yet of information about previously identified issues that could arise from the creation and deployment of these [AI] models.” The MIT group, and its many collaborators, sieved through mountains of peer-reviewed journal articles and preprint databases detailing the many and varied risks of AI unknowingly faced by the uninitiated. Of the 700-plus pitfalls identified, 76% related to the safety and robustness of AI systems, 73% to socioeconomic and environmental harms, 71% to discrimination and toxicity and 68% each to privacy and security and malicious actors and misuse. However, probably the greatest concern identified is that 90% of the risks inherent to an AI model become apparent only after it has been made available to the public, who unwittingly use it and face real dangers without being aware of what they are or how they happen: A mere 10% of risks are identified before deployment. As reported in the latest edition of the inestimable MIT Technology Review, the director of MIT FutureTech and one of the creators of the new database, Neil Thompson, commented: “These findings may have implications for how we evaluate AI, as we currently tend to focus on ensuring a model is safe before it is launched. What our database is saying is, the range of risks is substantial, not all of which can be checked ahead of time, Therefore, auditors, policymakers, and scientists at labs may want to monitor models after they are launched by regularly reviewing the risks they present post-deployment.” And then some – it should be mandatory! Valuable and pertinent though the AI Risk Repository is, it does not rank the potential dangers by the levels of risk they pose. That seems like a cop-out, but the lead author of the database, Peter Slattery, a postdoctoral associate at MIT FutureTech, stated:, “What we really wanted to do was to have a neutral and comprehensive database, and by neutral, I mean to take everything as presented and be very transparent about that.” The trouble is that such an approach immediately self-restricts the repository’s practical usefulness by not presenting the risks in league tables with the ones most to be avoided placed at the top of the list. What’s more, no guidance is given on how AI ingénues might defend themselves. The new database is a good idea and a good start but it will constantly need to be monitored, edited and expanded otherwise it may turn out to be an academic exercise of little practical application.

Turkcell is the latest telco to try out Ericsson’s Cloud RAN solution and in doing so is taking an approach to Open RAN that might become increasingly common, in that it is currently interested in checking out the network management tools and applications that have been developed for, but are not exclusive to, Open RAN deployments but not necessarily so fussed about multivendor options for the distributed units (DUs) and radios. “By adopting Cloud RAN technology, Turkcell aims to introduce new software features and incorporate an Open RAN approach into its service management and orchestration (SMO) framework,” noted the Swedish vendor in its announcement. “It will also be able to improve network management capabilities with rApps, for both cloud RAN and purpose-built deployments,” added the vendor, where rApps are software applications that run on a non-real time RAN Intelligent Controller (RIC). Prof. Dr. Vehbi Çağrı Güngör, chief network technologies officer at Turkcell, noted: "As one of the first examples in the world, the trial of the 5G Cloud RAN technology conducted with Ericsson will contribute to strengthening Turkcell’s strategy for cloudifying network operations. This initiative, which represents a new step in Turkcell’s technological leadership, will also support our transition to an Open RAN network strategy and enhance the advanced connectivity experience we offer in Türkiye.” 

Here at TelecomTV we’re boiling over with excitement about the multiple ongoing developments related to the development of quantum computers and, more particularly, the efforts underway to ensure quantum-safe networking. We noted last week that BT and IBM had plenty of positive things to say after the US Department of Commerce’s National Institute of Standards and Technology (NIST) issued the first three encryption algorithms for post-quantum cryptography (PQC) and they are not alone. Nokia, which has also been investing in the development of quantum-safe networking capabilities in recent years, published this blog in the wake of NIST’s announcement. “This is a momentous decision for the field of post-quantum cryptography (PQC),” noted Aritra Banerjee, a senior research scientist working on Nokia’s quantum-safe standardisation efforts, and Dimitris Schoinianakis, a senior security researcher at Nokia Bell Labs, who co-authored the blog. “NIST’s approval of these algorithms means they can now move to the different industry standardisation bodies, which will make them part of internet, network and data encryption standards that protect the world’s information. These PQC standards will be key to creating quantum-safe networks and products. The new algorithms will become powerful weapons in our growing quantum-security arsenal, defending us against the inevitable quantum computing attacks of the future,” they added. Check out the blog here. 

- The staff, TelecomTV