Holistic platforms, not one-offs, are key to AI success, report says

Development and adoption of AI must be done in a systematic and unified manner across managed service providers' (MSPs') departments and initiatives, according to ABI Research.

Using AI solely as a tactical response to problems or to exploit specific opportunities without creating a companywide vision and implementation strategy will limit artificial intelligence's huge potential.

"AI adoption is focused on narrow use cases whereby telcos are aiming to tackle specific pain points," wrote ABI Senior Analyst Don Alusha in response to emailed questions from FierceTelecom. "We argue that AI suppliers must enable an adoption roadmap wherein use case specific actions taken now yield short term desired outcomes, but also have an effect that radiates out for years to come towards a wider AI stratagem."

The firm found MSPs are using AI for customer services, network management, service assurance and cybersecurity. The problem is that AI now is employed narrowly in numerous uses cases. Each employs its own solutions partners, frameworks and roadmaps.

RELATED: IBM unveils tools to detect bias in artificial intelligence

The MSPs must evolve beyond this parochial world. "They see huge potential in adopting AI, but much of the adoption is tweaked and fine-tuned in line with specific use cases associated with individual products or departments," Alusha wrote. "For example, we found that intelligent user interfaces, sales and marketing are use case that currently dominate in telecoms."

Alusha said that some organizations are starting to consider the way forward. "We would say that for MSPs, this is an ongoing journey and they are increasingly cognizant of the fact that a unified, wider AI strategy is needed. We expect to see more discussion around this in the coming months. Telecom vendors have a key role to play here and some, such as Ericsson and Huawei, are investing with an eye to enable a more widespread, harmonized approach."

The firm pointed to cases in which a holistic approach already is being used. AT&T's Acumos and Telefónica's LUCA projects are implementing AI across many platforms. This, Alusha suggests, needs to be the accepted approach for AI to reach its potential.

Putting a stable and companywide structure in place will enable AI to become a foundational element that can help the company thrive. MSPs, Alusha wrote, "should overcome divisions on two strands to fully capitalize on benefits of AI. One is digital divisions such as OSS/BSS assets and their networks. The other is strategic in the form of new ethical frameworks, governance systems and most importantly, the necessary human capital. All are important aspects, but the human capital is key. They must create the right culture to integrate AI into their DNA to create a workable human plus AI business model."

There is much concern about the power of AI, especially when it is synergistically combined with big data, the IoT and other potent technologies. To many, the sheer power of AI means that special precautions and steps must be taken. "The ethical view is important, but given that telecoms is still at an embryonic stage in adopting AI, there is very little discussion on that front," Alusha wrote. "It is however a very mature discussion in the technology domain for Facebook, Google, Amazon and others."