Skip to the content
Australians for AI Safety

AI discrimination: what it is, and how experts rate it

An AI system treats people differently on the basis of race, gender, age or disability, usually because it learned from records of decisions that already did. Unlike a biased individual, a biased system applies the same error to everyone it touches, at speed, and without anyone being able to point to the moment it happened.

  • Job screening
  • Loan and insurance decisions
  • Learned from records that already discriminated

Unfair discrimination and misrepresentation. Unequal treatment of individuals or groups by AI, often based on race, gender, or other sensitive characteristics, resulting in unfair outcomes and representation of those groups.

Definition and category 1.1 from the MIT AI Risk Repository, used under CC BY 4.0.

How experts rate it

9.1%

With mitigations — 5.5%The difference between the two estimates — 3.6 points

That threshold is the wrong lens for this harm. AI discrimination is widespread, serious and individual, so a low number here does not mean a small problem.

Mean probability of a catastrophic outcome by 2030 — more than 1 million deaths, or more than USD 100 billion in financial loss, or comparable civilisation-scale intangible damage. Across 272 experts who took part. How this was measured

The study tabulates 5 of its 23 figures; this one was read from MIT's published chart on 18 August 2026.

Ranks 22 of 23 rated risks. This study published no interval for it, so the position of neighbouring risks is not meaningful. Experts put the difference at 3.6 percentage points.

Rank 22 of 23 by expert-rated severity, at 9.1%. See all 23 ranked

Evidence

Nothing here yet. Know of an incident, study or report? Add it.

What's been called for

Nothing here yet. Know of something that's been called for? Add it.

What politicians have said

  • “Concerns about accuracy, bias, client confidentiality and broader ethical concerns in relation to the use of AI cannot be lightly dismissed”

    Michelle Rowland · · Australian Law Forum Speech

    AI discrimination

See all statements

What Australians say

    Nothing here yet. If this matters to you, say why in a sentence.

    What people ask

    Are you saying that all AI systems are dangerous?

    No. Today's LLMs are unlikely to pose catastrophic or existential risks. While there are valid ethical concerns, today's AI systems offer important economic opportunities while being unlikely to pose extreme safety risks. However, evaluations by leading labs and third parties show an increasing likelihood that LLMs will soon pose a range of risks, like being able to assist wrongdoers build chemical, biological, radiological and nuclear weapons, conduct sophisticated cyber attacks at scale, or even escape human control. Given AI scaling laws, the need to prepare for highly capable AI systems that could be dangerous has become pressing.

    See all FAQs

    Think Australia should act?

    Your MP and senators decide what Australia does about AI. A few lines in your own words about AI discrimination is enough to tell them it matters to you.

    Write to your representative