Artificial intelligence (AI) might hold the potential to be fairer than humans. Ipsos iSay team reveals the findings from a global Ipsos survey on trust in AI and its impartiality.
Do we trust AI more than ourselves when it comes to avoiding discrimination?
Discrimination and bias are persistent issues in our society. But could AI offer a more equitable solution? According to a global Ipsos survey, 54% of people trust AI to avoid discrimination or bias against any group of people, while only 45% trust humans to do the same.
These results are intriguing, especially considering that AI is created and trained by humans. However, the perception is that AI, being data and algorithm-driven, can bypass the inherent prejudices and biases of humans.
People are less trustworthy than AI: all generations think the same!
That's 9 percentage points less trust in humans than in AI not to discriminate or show bias towards any group of people! Surprisingly enough, the trend is similar across 4 generations, from Gen Z to Baby Boomers, all trust AI more than humans to be impartial.
This reflects a lack of confidence in our own ability to be fair and just. AI, on the other hand, is viewed as an impartial tool that could help level the playing field.
But is this truly the case? In reality, AI can also be biased if the data it is trained on is unrepresentative or tainted with prejudice. Therefore, it is crucial for AI developers to be aware of these risks and strive to mitigate them.
What do you think: Can AI be fairer than humans? At Ipsos iSay, we want to hear your opinion. Participate in our surveys and share your thoughts on this significant topic.
Source: Ipsos survey conducted among 23,685 adults under 75 years old from 32 countries between April 19 and May 3, 2024.