Data Analysis
Bias In A.I.

Healthcare, the justice system, even playful facial-recognition based apps: the number of
instances in which researchers have identified how bias in A.I. is harmful keeps growing.

As experience designers, Thinkwell wants everyone who experiences the things we design and create to be welcomed and have a supportive, equitable experience. Thinkwell’s Trend Report data set analyzes this issue more deeply: are respondents concerned about A.I. bias and what are their experiences with it, from the perspectives of race, age, and gender.

Race

  • Concerned about the potential for bias in A.I. technologies

  • Experienced a form of bias by an A.I. technology (personally or someone they know)

Thinkwell's 6th Annual Guest Experience Trend Report survey supports the growing body of data of bias in A.I. technologies. According to our survey, people of color have more frequently experienced bias by an A.I.

Asian respondents are the most concerned about the potential for bias in A.I. technologies, Black/African American respondents are the second most concerned, with white respondents being the least concerned. The level of concern of different demographic groups roughly correlates with their likelihood to experience a form of bias by an A.I.

Age

  • Concerned about the potential for bias in A.I. technologies

  • Experienced a form of bias by an A.I. technology (personally or someone they know)

Younger respondents are more concerned about A.I. bias than older respondents.

Moreover, the percentage of respondents who have experienced a form of bias by an A.I. technology has plummeted from 32% to 11% as participants’ age increases. As there is no study showing that A.I. algorithms are pro-seniors (on the contrary, studies have proven certain A.I. technologies to be ageist), the differences in the level of digital literacy and attachment to technology among younger and older generations may have contributed to this dramatic drop.

Gender

  • Additionally, one non-binary respondent reported concern.

    Concerned about the potential for bias in A.I. technologies

  • Additionally, one non-binary respondent reported no experience.

    Experienced a form of bias by an A.I. technology (personally or someone they know)

Both men and women are equally concerned about bias in A.I. technologies.

While men were more likely to report with certainty actual instances of bias occurring to them, it’s possible the share of “unsure” responses indicates that women may cast more doubt about their own experiences and underreport instances of bias relative to men, which could bring the rates between groups somewhat closer.

Conclusion

Bias in A.I. is a prevalent concern for guests of all ages, genders, and racial backgrounds. For this reason, Thinkwell is convinced of the significant need and value in recognizing the challenges and addressing them through thoughtful and proactively anti-bias A.I. design.

Exploring FURTHER >

Continue exploring Thinkwell's 6th Annual Guest Experience Trend Report subjects
by clicking on the deep dive reports and data analyses below (with more to come in
the weeks ahead)!

  • Thinkwell's 6th Annual Guest Experience Trend Report

    How artificial intelligence can affect, adapt, and improve the guest experience.

    Return
  • Data Analysis: Generation Gaps

    To what degree does age determine attitudes towards A.I. technologies?

    Read Article

Interested In Learning More?

Reach out through the form below and let’s chat. PDF version of the report can be downloaded here.

Download

Thinkwell's 6th Annual
Guest Experience Trend Report

Artificial Intelligence

©2022 Thinkwell Group, Inc.