Skip to main content Skip to footer

The age of artificial intelligence (AI) is here, and countries in the Asia Pacific (APAC) and Middle East zones have set their eyes on the prize. Start-up activity around AI is booming, and the region is set to overtake the rest of the world in AI spending over the next three years — reaching $15 billion by 2022, according to IDC.

Against this backdrop, we surveyed 590 senior executives across the region in late 2019 to understand their companies’ AI plans and actions. Our goal was to learn how APAC and Middle East businesses are employing AI, how this emerging technology is impacting business and how they are overcoming challenges to reap value from machine intelligence across various functional areas. (To learn more about the study, including its methodology, see our white paper, “How Companies Can Move AI from Labs to the Business Core.”)

While our research was conducted before the outbreak of COVID-19, we believe the pandemic only underscores the urgency around implementing AI — and doing so thoughtfully. Advanced forms of AI are critical to our understanding and treatment of infectious disease. Machine learning (ML) and deep learning are helping infectious disease scientists more quickly sequence and model treatment therapies and vaccinations, accelerating time-to-market and limiting unintended consequences. They are also helping researchers more effectively forecast and foretell the contours of pandemics. These capabilities are applicable to every business in every industry.

Despite significant progress, AI’s impact in APAC and the Middle East is today limited by the fact that businesses remain in experimentation mode (see Figure below). If this is a cloud, there is indeed a silver lining: organisations in the region can learn from the hard-won experience of businesses elsewhere. With this in mind, we prescribe the following procedures to shift AI from experimental mode to production mode.

  Formulate a strategy that lets business value guide the choice of AI opportunities.
  • Avoid common errors. AI endeavours can be seen as mere technological projects, rather than steps in building a business value proposition for the technology. This attitude undermines AI’s potential upside. Similarly, businesses must avoid the ivory tower mentality, where AI enjoys a higher priority over the digital transformation of the enterprise, in which everyone is a stakeholder.
  • Look out for execution blind spots. Ignoring blind spots is an invitation for unwanted outcomes. To avoid them, execute broadly across the business, looking for as many opportunities as possible to give yourself multiple chances at success. Also, avoid big bets that can divert emphasis from the enterprise’s digitizing efforts.
  • Develop a robust strategy. An AI strategy should be a microcosm of the broader business strategy for digital transformation. This means incorporating key business metrics to ensure that AI helps the organisation maintain an unwavering focus on business outcomes.
  • Combine a business impact criterion with technological capabilities. While approaching any AI use case, organisations should ask how that project helps their business. An idea may seem like a sure-fire success up to the pilot stage — but fail to deliver the desired results. Similarly, businesses often get carried away trying to perfect the technology instead of driving transformation. To this end, our critical recommendation is to start with a business case based on a sound hypothesis.
  • Rethink the ways of business first; then spot the resulting opportunities. Wherever necessary, businesses must put in place new support structures before exploring AI opportunities. This begins by reimagining how work gets done now that machines can do more and more of it, and viewing data through a holistic lens. Importantly, it means building trust for ML. At a mining company in Australia, worker injuries and deaths were increasing. We stepped in to craft an AI-based solution capable of predicting risk events, combining the power of ML, text analytics and natural language processing. Result: a drop in injuries that enhanced the brand’s reputation.
2    Combine business and behavioural insights to build human-centric AI.
  • Embed human-centricity up front. Human-centricity is critical for balancing ambition with machine resilience. This means an early and unwavering focus on AI literacy, re-skilling, up-skilling and retooling.
  • Factor in behavioural insights. Today, behavioural science is more important than ever for businesses. AI’s transformative impact on businesses and industries means it needs to be a valued part of the organisation. To this end, businesses can fuse learnings from sociologists and anthropologists with their business strategy inputs.
3    Build a data-driven organization.
  • Embed data analytics to drive decision-making. AI is only as good as the data it accesses. For businesses, it is not only critical to get a holistic view of data, but also to use data analytics to drive decision-making across the organisation.
  • Focus on compliance with data privacy laws. Compliance with data privacy laws is critical to the trust established among businesses, governments and customers. As their AI deployments evolve, organisations must simultaneously safeguard the privacy rights of everyone concerned.
4    Deploy AI tools and algorithms in line with the maturity curve.

At a given point in time, any two organisations will confront different problems and challenges. Achieving AI maturity involves moving along the maturity curve iteratively, adopting the necessary tools and creating new algorithms. Appropriate options will be defined by where organisations find themselves on the digital maturity curve.

5    Pave the way for creating an AI-centred culture.

The cultural rewiring of a business begins at the top and should be approached with the end goal of becoming a data-driven organisation in which human creativity thrives around AI. To that end:

  • Inculcate cross-functional cooperation. To realise AI’s transformative power, businesses need a cross-functional team and a structured approach to identifying opportunities for process improvements.
  • Overcome risk aversion. As businesses move beyond experimentation, they will need an environment that encourages creativity and a risk-taking attitude. To make this happen, we recommend setting up an AI office or centre of excellence to oversee AI projects from ideation to production. Small, multi-skilled teams are critical. AI success depends on combining knowledge from business functions, processes, data and technology.
  • Close the learning loop. Bridging the gap between learning from AI experiments, which happens simultaneously across the business, and the areas in which the lessons can be applied is important for advancing capabilities. An organisational mechanism that can oversee the experiments and update the broad approach accordingly can go a long way in upgrading capabilities.
  • Keep on communicating. A strong culture thrives on communication. AI deployments might succeed or fail, but communication keeps people going. It is critical for any corporate AI effort to keep all channels open for sharing knowledge and information.
6    Embrace ethics and governance.

Embed ethics up front to build responsible AI applications. As AI is expected to be pervasive, touching every aspect of the organisation, building ethics into the fabric of AI technology is essential. Companies must embed a focus on ethics from the initial development stages, and must never relinquish that focus, even as AI apps themselves evolve and learn. Governance models are critical in ensuring that ethical aspects of AI don’t get overlooked. AI’s current limitations in the form of built-in biases and an inability to handle complex situations are well documented. As businesses advance their AI efforts, ethics and governance become more and more critical to their deployments.

7    Employ external partnerships as AI levers.
  • Address talent gaps. Choose external vendors that are competent to augment the business’s AI capabilities — vendors that have a workforce with AI-ready skills to engage an effective partnership.
  • Leverage partnerships to access ever-advancing AI technologies. How can organisations better prepare for rapid AI experimentation? Securing ready access to new technologies and techniques is an important first step. Many AI efforts get bogged down in lengthy technology procurement processes. Businesses need to be sure they have an open cloud environment to experiment with machine data. Better yet, they should create a robust set of partnerships that provide access to continuously advancing AI technologies.

To learn more, visit the AI section of our website or contact us.