- 53% of enterprises have now moved beyond AI pilots, a marked increase from 36% in 2017 based on Capgemini’s previous study of AI adoption.
- U.S., Chinese and South Korean governments have stepped up the use of AI technologies to contain the spread of Covid-19 and to handle citizens’ concerns.
- Life Sciences, Retail, automotive, consumer products and telecommunications are the top five industries leading A.I. adoption today.
These and many other insights are from Capgemini Research Institute’s research of enterprises’ AI pilot, production and spending during the pandemic provide insights into what differentiates AI-at-scale leaders versus organizations struggling to get results. The AI-powered enterprise: Unlocking the potential of AI at scale is based on 950 interviews with AI leaders in enterprises with annual revenues between $1B to $25B, distributed across 11 nations. The study was completed in March and April 2020. For additional details on the methodology, please see pages 31 and 32. Capgemini has made the report available for free download here (PDF, 40 pp, no opt-in).
Why AI-at-scale Leaders Are Getting Results
The study’s research team defines AI-at-scale leaders as those enterprises that have launched multiple AI applications across numerous departments and are seeing results from their use. 13% of enterprises have achieved AI-at-scale leadership and 72% have been unable to deploy even a single application. AI-at-scale leaders excel at developing and launching apps despite the pandemic and resulting economic downturn, showing formidable resilience and organizational strength in the process.
Here are the key insights from the study:
- The top 13% of enterprises are AI-at-scale leaders, outpacing their peers at moving multiple AI pilots into production. Capgemini’s research team finds that one of the main reasons why AI-at-scale leaders progress faster than their peers is their ability to create tight collaboration between business and IT, creating shared accountability for results. Nearly half of the enterprises (46%) are running AI pilots and Proof of Concepts (POCs) and have not yet deployed an app into production. POCs can be problematic as every department often agrees for the need to do one, with no department wanting the responsibility if it succeeds or fails.
- Life Sciences leads all AI-at-scale enterprises with the highest levels of AI adoption across multiple business teams. 27% of all AI-at-scale enterprises are from the Life Sciences industry, followed by Retail (21%), Consumer Products and Automotive (both 17%) and Telecom (14%). On average, pharma companies spend 17% of their revenue on R&D, among the highest percentage across industries. They have also made significant AI bets in areas such as drug development, R&D and diagnosis. AI adoption shows the greatest potential for growth in telecom enterprises today, with 57% having deployed a few use cases in production on a limited scale. Interestingly, Public/Government and Consumer Products are the two industries where pilots and POCs stall or stop. 59% or nearly six in ten AI pilots and POCs in Public/Government run the risk of stalling out and not moving to production.
- 78% of the AI-at-scale leaders continue to make progress on their A.I. initiatives at the same pace as before the Covid-19 pandemic. Over one in five (21%) of the AI-at-scale enterprises have increased the speed of their deployments. Conversely, 58% of struggling organizations have either pulled investments from AI initiatives with low potential impact due to high business uncertainty or suspended all AI efforts.
- 97% of the AI-at-scale leaders have seen quantifiable benefits from their deployments, compared to 64% of the struggling organizations. Also, AI-at-scale leaders are much more likely to have achieved benefits that met or exceeded their expectations (94% compared to 59% of the struggling organizations).
- 79% of AI-at-scale leaders have seen more than a 25% increase in sales of traditional products and services this year. Leaders are using AI to improve every phase of customer lifecycles, evidenced by 62% of them seeing at least a 25% decrease in the number of customer complaints. Learning how to launch AI apps successfully delivers solid results in other areas of an enterprise as well. 71% of the AI-at-scale leaders see a 25% reduction in security threats as a result of improved cybersecurity.
- AI-at-scale leaders concentrate on initial large-scale deployments that provide every functional area an opportunity to take ownership of outcomes and the success of the project. 60% or more of the AI-at-scale enterprises have use cases in functions spanning from customer to risk and compliance to people. Taking an end-to-end approach to prioritizing user cases not only benefits the teams of that particular business function but also ensures that the various business units across the organization become enthusiastic stakeholders in the AI development, launch and use. The following are the top-most implemented use cases AI-at-scale leaders have deployed in different functions.
- Multinational efforts between China, South Korea and the U.S. to contain the spread of Covid-19 and handle citizen’s concerns using AI are also delivering results. Contact tracing pilots and POCs are moving quickly to production testing, capitalizing on algorithms used to analyze overall data and identify first-, second- and third-degree contact tracing. The U.S. Department of Defense is using AI, machine learning and data visualization tools to spot potential Covid-19 hotspots. Several hospitals in the U.S. are using AI algorithms to predict which patients will become critically ill, up to 40 hours before a life-threatening event could occur. The State of Texas is using AI-driven chatbots to answer questions from unemployed residents in need of benefits as well according to the study.