Research: Mistrust of AI runs deep, but many remain optimistic for its future


Research: Mistrust of AI runs deep, but many remain optimistic for its future

We surveyed knowledge workers to understand sentiment about AI—from its trustworthiness to its future as an enterprise essential.

Jon Cohen

September 25, 2021 | 4 min read


In the business world, there is notable tension between the importance of artificial intelligence (AI) in enterprise systems and the technology’s perceived association with bias. 

We recently surveyed 1 533 knowledge workers who work at medium and large-sized organizations to better understand sentiment about AI. The disconnect that we uncovered tells a story of a pervasive technology that people believe has both potential and significant room for growth. 

Here’s what we learned about how people view AI’s capabilities, trustworthiness, and future in enterprise business. 

For the majority of people, AI bias is a deep concern

A full 61% of knowledge workers said that the data that feeds AI is biased. That belief is higher among people of color; 75% of Hispanics/Latinos, 71% of Blacks, and 62% of Asians cited the existence of biased data in AI-powered systems, compared to 58% of white respondents. 

Graph showing the percentage of belief that the data that feeds AI is biased, by race/ethnicity

That such a significant number of people acknowledge bias in AI systems is in keeping with broader, industry-wide concern. In the business world, when algorithms are trained using biased data, there are major ramifications for companies’ products, user experience, and reputation. From recruiting tools and credit card application algorithms to UX features like automated photo-cropping, when AI is fueled by biased data, it can lead to bad predictions, discrimination, and the reinforcement of racist and sexist systems.

Of course, data bias doesn’t occur in a vacuum—and where there’s data bias, there are likely other issues at play. For instance, 48% of workers believe that AI algorithms are not written by engineers with diverse backgrounds. 

Homogenous AI teams are something many industry leaders have cited as the technology’s primary obstacle and the root of its bias-based flaws. “The machine we teach can only be as good as its teacher,” said Jing Huang, director of engineering-machine learning here at Momentive. “If there isn’t diverse representation in the data scientists and engineers who are building the algorithm and teaching the machine to learn, you won’t have a generation of artificial intelligence that is representative of our human society.”

The issue of inclusion in the development of AI technology applies not only to race, but to gender representation as well. In our study, 49% of people who identify as male claimed to have expert-level knowledge of AI, while only 28% of those identifying as female said the same—a sign, perhaps, that fewer female workers are immersed in the technology in their day-to-day work.    

Yet AI is the future of enterprise

Significantly, despite any concerns workers may have about biased data or a lack of inclusive AI teams, the technology is still an enterprise essential. A majority of people (86%) said that AI is “important” or “very important” when their organization is selecting enterprise technology to purchase. Executive decision-makers were even more likely to say this, with 96% of respondents who hold C-suite, president, or owner positions saying that it’s “important” or “very important”. 

And AI technology isn’t limited to one type of workflow or area of business, but rather interwoven throughout respondents’ companies. The use of AI-driven software spans departments, with the highest use falling to IT (54%), marketing and sales (44%), and finance, procurement, and accounting (41%). 

All signs indicate that companies believe their future lies with AI—and that AI could make or break which systems, software, and partnerships are deemed crucial for business.

graph displaying the importance of AI when selecting enterprise technology to purchase

Despite misgivings, AI optimism persists

Perhaps because of AI’s integration into the business systems used every day, people ultimately believe in the technology’s benefits and opportunities. Among our respondents, 39% believe AI helps humans become more effective overall, and 40% believe that it helps spot trends that human data analysts could overlook. The latter sentiment is even higher among older adults; 61% of people 60 years or older believe AI can augment the abilities of human analysts. 

Graph showing AI optimism about jobs, based on age

One area of business where AI’s use is often most apparent is customer service, where AI-powered communication is designed to enable 24/7 service and improve relationship management. People are taking notice, with 88% saying that AI has the potential to improve customer service. And some workers believe other areas of business could potentially benefit from advances in AI. Nearly a quarter say that AI will generate millions of new jobs, with slightly more optimism (31%) coming from those ages 30-39. 

As for truly objective AI, people remain hopeful: 73% believe that AI is fully capable of making unbiased decisions on behalf of humans.

Inga Vailionis leads customer insights on the brand marketing team at Momentive.


This Momentive study was conducted in May and August 2021. We surveyed 533 knowledge workers, aged 24 and older, who live in the U.S. The sample was balanced by age and race, among other demographic variables, according to the U.S. Census.


Read more articles by Momentive

Stay up to date on the trends that are reshaping industries, the innovations that are bringing humanity and technology closer together, and more.


Momentive study: half of Americans would consider an EV

Van Westendorp Pricing Analysis shows consumers willing to pay $20,000 - $30,000 for an EV

Dec 1 • 4 min read

Read more


Momentive study: teens flock to immersive worlds

Teen gamers choose Roblox as their favorite “game,” and show interest in VR and the metaverse.

Oct 21 • 5 min read

Read more