Questions people ask · Safety and trust

Is AI biased against people like me?

It can be. AI learns from data, and data over-represents some languages, cultures and viewpoints while under-representing others. If an AI system screens job applications or recommends treatments, hidden assumptions could disadvantage people who don't fit the patterns it learned. That is why testing matters. In the UAE, the Central Bank now expects the financial firms it licenses to test their AI for bias at least once a year and to let customers ask for human review of AI decisions [REG7]. Arabic-first models aim to reduce cultural gaps [QA3, UAE11]. If you think an automated decision treated you unfairly, ask for the reasons and for a human review. → Fears vs Reality: Is AI biased against people like me?

A fuller, step-by-step guide

Articles on this question

All questions