skip to main content


This report details the challenges and opportunities facing IT service providers in applying AI technologies and techniques in businesses in a meaningful and useful way.


  • Acquisition of startups is dominated by technology giants (Alphabet, Apple, Facebook, Amazon, Baidu, Alibaba). IT services companies such as IBM and Accenture have been acquiring too, but are trailing the tech houses.
  • The AI talent crisis is likely to get worse before it gets better. The bigger tech giants can attract top talent from a limited pool of AI academics, data scientists, and skilled practitioners. This makes building AI competency difficult for everyone else.

Features and Benefits

  • Identifies opportunities and challenges related to AI implementation
  • Illustrates use cases from a variety of IT services providers

Key questions answered

  • How can AI be implemented in a large enterprise or organization?
  • Is digital transformation a prerequisite for artificial intelligence?

Table of contents


  • Catalyst
  • Ovum view
  • Recommendations

AI – almost 70 years young

  • Five critical developments have triggered a resurgence in machine learning
  • Artificial neural networks (ANNs) underpin machine learning and deep learning
  • Where deep learning makes sufficient progress, it can be productized, or at least road tested
  • We are at the beginning with AI; it’s just the dawn of day one

Implementing AI in large enterprises and organizations

  • Large enterprises and organizations need navigators and design expertise
  • Proven methods and productized components are a good launch pad into AI
  • Open source for everyone, differentiation will be in the solution
  • Chatbots, image recognition, and analytics are the applications seeing the most activity
  • Rules-based automation is not AI, but it is a gateway to AI

Challenges and opportunities in implementing AI

  • The adequacy of black-box style AI is questionable when accountability matters
  • A backlash grounded in ethics could stall progress
  • Use cases where accountability matters less are easier to progress
  • The acceptance criteria for AI will be subject to broader considerations
  • The skills shortage is likely to get worse before it gets better

Acquisition of AI startups is rampant

  • There is a range of buying persona in a white-hot market

Market vision for AI

  • The exploratory phase will give way to industrialization
  • Digital transformation is the critical enabler of making artificial intelligence work


  • Methodology
  • Further reading
  • Author