Editor's note: Joe Rydholm can be reached at joe@quirks.com.

Our cover story this issue chronicles one research agency’s process for deciding how to move forward with AI and it’s no doubt a scenario being played out all across the business world. While more and more researchers – agency and in-house – are dipping their toes in the AI water, many more, especially those on the client side, are hamstrung by a range of factors, not the least of which is dearth of internal direction from the people running their organizations, whether it’s the IT and legal departments worried about data privacy and protecting corporate IP or C-suiters still trying to wrap their heads around what AI is and what it could do for (and to!) their businesses.

A recent study of executives and AI leaders conducted for Teradata by research firm NewtonX echoed that sense of uncertainty, finding a general lack of confidence in AI strategies and nervousness about the readiness of AI outputs.

While 89 percent of enterprise executives believe AI is necessary to stay competitive, only 56 percent say their companies have a clear AI strategy and only 28 percent see their AI strategy as closely aligned with and supporting broader business objectives. 

Company leaders know that AI has to be introduced and implemented ASAP across the whole enterprise but so far, the survey data found, most successful AI implementations are happening at the departmental level: just 12 percent have deployed AI solutions company-wide, while 39 percent have implemented AI in select departments.

Among those surveyed, the focus seems to be on using AI to increase productivity/cut costs and improve the customer experience but there are worries that AI could potentially damage a company’s relationship with its customers if something goes wrong – obviously a reflection of it being early days with this technology and as a result, we have no real sense of how bad things could be if disaster were to strike.

Trust is key

At the core of all of this, of course, is trust. Trust in the veracity of the data AI generates. Trust in the validity/lack of inherent bias of the information AI is using to produce its outputs. 

As one study participant said, “…we want to be very clear with our customers what data has been used to train the models,” noting that it can be easy to introduce bias into the models by choosing the wrong training sets. Another said, “…master data management is not glamorous but … if you’re basing everything off the data and the data is flawed, then you’ve got a problem.”

More than half (57 percent) of executives surveyed said they are concerned about how AI missteps could impact customer satisfaction, company reputation or both, noting that there needs to be greater cohesiveness between AI and business planning for it to be successful.

Even with internal projects, 63 percent of executives surveyed report using a mix of closed and public data sets, while only 29 percent rely exclusively on closed data sets.

Barriers to scaling AI projects effectively include: scarcity of AI technical talent (39 percent); lack of budget required to scale AI projects (34 percent); difficulty in measuring business impact (32 percent); and insufficient technology infrastructure (32 percent).

About half of executives surveyed have successfully leveraged AI to enhance employee productivity and collaboration (54 percent) and support decision-making (50 percent), yet only a third have used AI for product development (30 percent) or sales and revenue forecasting (30 percent). 

Immense pressure

And, perhaps unsurprisingly, the respondents expressed a feeling of being behind on their AI adoption. With so much press coverage and discussion about AI, leaders must surely feel immense pressure to get with the cool kids and scale up their AI efforts in a hurry. While 73 percent of those surveyed see their companies as early adopters with many technologies, 60 percent said their level of AI adoption is merely “on par” with their competitors. Only 27 percent see themselves as leaders of AI adoption in their industries. 

The survey was distributed in the U.S., Europe, the U.K. and Asia and polled C-suite executives and AI decision-makers in companies with at least 1,000 employees and more than $750 million in annual revenues. The survey reached ~300 AI-relevant executives, from companies like Nike, P&G, Hermes Paris, Allianz Partners, Prudential Financial, Honeywell and Novartis, with about half of the respondents located in the U.S.