To capture the pulse of the buying public, InRule Technology® contracted Boston PR firm PAN Communications to survey consumer attitudes toward AI. Key takeaways:
Consumers are wary of AI/ML, especially when used to make decisions affecting themselves…
- 43% of respondents stated that they have trust in artificial intelligence/machine learning (AI/ML) technology.
- 35% of respondents believe AI/ML makes decisions quicker than humans.
Level of trust in AI/ML increases with the level of education…
- 37% of high school graduates trust AI/ML technology.
- 48% of those with an associate’s or bachelor’s degree.
- 57% of those with a master’s degree.
Level of trust decreases with age…
- 41% of millennials trust AI/ML.
- 28% of boomers trust AI/ML.
Consumers are not confident in their understanding of AI/ML…
- 81% of respondents report having an average to below average understanding of how AI/ML works.
- 27% stated they have little to no understanding of bias within an AI/ML algorithm.
Consumers expect more transparency and less harmful bias in AI/ML…
- Nearly 40% stated that their perception of AI/ML would improve if companies were more explicit about how, where and when they use it.
- 57% of respondents think it is important for companies to proactively reduce bias within their AI/ML technology.
- Importance increases with education – 31% of high school graduates found it extremely important for companies to take proactive steps to reduce bias, 42% of those with a master’s degree and 48% Ph.D. holders
- 77% said they would stop using a company’s products or services if they learned their AI/ML technology was making biased decisions.
- 53% desired more government regulation over AI/ML technology usage.
While appreciative of its speed and helpful capabilities such as text bots, the majority of the buying public harbors various degrees of skepticism regarding its fairness in making decisions. InRule offers organizations the confidence of machine learning and decisioning tools equipped with the industry’s best-in-class bias detection.
InRule® Machine Learning delivers the why behind every prediction and decision, as well as the impact of each predictive component. ML model visualizations and unparalleled explainability empower users to swiftly assess predictive models for harmful bias and prevent undesired outcomes. Other bias-detection systems only evaluate for bias by averaging values across an entire model. Whereas the InRule high-capacity clustering engine assesses the deepest subsets of a model based on the similarity of each prediction, ensuring each model operates fairly.
Ensure your machine learning models are explainable and your corporate reputation stays well protected from harmful bias. Experience InRule Machine Learning and the full suite of no-code automation tools of the InRule® Intelligence Automation Platform. Request a free demo or a 30-day trial today!
