What 350+ IT decision makers had to say.
More than half of IT leaders (54%) say they’re very or extremely concerned about AI bias, up from 42% in 2019.
What are They Worried About?
56% Loss of customer trust
50% Compromised brand reputation
43% Increased regulatory scrutiny
Negative Impacts of AI bias
AI bias has already affected organizations across the globe. 36% have suffered from AI bias in one or more algorithms—with damaging impacts:
62% of organizations lost revenue
61% of organizations lost customers
43% of organizations lost employees
Almost nine in ten (87%) organizations are taking actionable steps to detect and prevent indirect bias.
This detection and prevention effort is far stronger in the U.S. (88%) than the U.K. (29%).
How Organizations Are Battling AI Bias
69% Data quality checks
51% Training employees on AI bias
51% Hiring an AI bias/ethics expert
50% Measuring AI decision-making factors
Ways to Address AI Bias in the Future
The majority of companies (83%) are using either third-party systems or a combination of both in-house and third-party systems to help detect AI bias. But there’s still progress to be made: nearly one in five companies not using a third-party system have no requirements for checking bias, which means it’s still likely to occur.
On the positive side, most organizations know this is a problem. 84% are planning to invest more in the coming 12 months.
Where are They Focusing Their Efforts?
54% will increase resources to support model governance
52% will hire internal personnel to manage AI trust
51% will use more sophisticated AI systems
50% will use more explainable AI systems
41% will appoint a third-party vendor to manage AI trust
Download the full report for more insights and start getting ahead of AI bias.