Companies see value, challenges in AI

March 8, 2019
Jason Sparapani.

Companies today regard AI as more than just hype, with 70 percent of survey respondents saying they’ve adopted or plan to adopt the technology, and just 1 percent saying it will have no value for their companies.

But the optimism buoying AI has a flipside, with a majority saying there’s a shortage of internal expertise, and low-quality data and a high level of investment also cited as challenges.

Respondents to the survey, by MIT Technology Review Insights, are business professionals working at midsize or large organizations worldwide. The survey is ongoing, with 396 responding as of March 8. Nearly half of respondents, 46 percent, are from North America, 28 percent are from Europe, and 18 percent from the Asia-Pacific, which consists of Asia plus Australia and New Zealand. All values are approximate.

This analysis primarily focuses on North America, Europe, and the Asia-Pacific because the other two regions surveyed, Latin America and the Middle East and Africa, are not at this time represented by statistically significant sample sizes.

Untapped value

Most companies using AI are investing in the technology on a case-by-case basis, at 42 percent, the survey found. This may indicate that at more companies today, individual business problems are driving the decision to use AI. Companies with a centralized plan to roll out AI is nearly half that, at 22 percent, and 6 percent of companies have incorporated AI into their operations.

AI adoption was the same in North America and Europe, with 71 percent of respondents in each region investing in AI case by case, launching AI strategies, or incorporating AI into their operations. Adoption was slightly higher in the Asia-Pacific, at 76 percent.

In North America, 19 percent of respondents said they believe AI will have value for their companies but have not yet adopted it, while 8 percent have not determined whether AI could have value. Europe showed a nearly identical dynamic, with 20 percent optimistic about AI but not yet taking it on, and 8 percent having made no determination about the technology’s value. In the Asia-Pacific, the divide was 19 percent versus 4 percent.

As for where companies are using AI the most, 32 percent are applying it to product or service development, 22 percent are using it in research and development, and 10 percent have incorporated AI into customer service. Smaller percentages of respondents are using the technology in other areas of the business, with 6 percent applying AI in sales and marketing, 5 percent in their supply-chain operations, and 2 percent in finance.

Trouble with AI

As with any new technology, AI poses numerous challenges for companies. Respondents to the survey were asked to choose three they considered the most pressing. The number one difficulty for AI adopters was a dearth of talent, with 61 percent of respondents saying they had trouble finding AI expertise and experience among their workforces. Nearly half, 48 percent, cited low-quality data for AI as a problem, and 37 percent said the high level of investment required was a challenge.

Other problems survey respondents pointed to were regulatory issues such as data protection laws, with 27 percent; difficulty making a business case for AI, with 24 percent; internal resistance to change, with 20 percent; slow decision-making among executives, with 19 percent; and labor union or workforce issues, with 2 percent.

In North America, 62 percent saw a shortage of talent as a problem. In contrast, 14 percent said slow decision-making in the C-suite was a challenge. This could be a sign that companies rolling out AI technologies have a fair supply of people who can think strategically about AI and few doers who can work with the technology. In Europe, 56 percent saw the talent shortage as problematic in deploying AI, while 19 percent blamed C-level foot-dragging. In the Asia-Pacific, it was 71 percent versus 25 percent.

This report was written by Jason Sparapani

Top photo credit: