Skip to main content
MIT AI Risk Initiative
BackBias and discrimination (bias in training datasets)
Home/Risks/G'sell (2024)/Bias and discrimination (bias in training datasets)
Misinformation and disinformation
Bias and discrimination (value embedding…

Bias and discrimination (bias in training datasets)

Regulating under Uncertainty: Governance Options for Generative AI

G'sell (2024)

SourceDOI
Sub-category
Risk Domain
1Discrimination & Toxicity
1.1Unfair 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 unfair representation of those groups.

"AI experts consider training data to be the most salient source of bias in generative AI models. For example, GPT- 2’s training data comes from outbound links from Reddit, a social network often criticized for hosting anti-feminist content.351 As a result, AI models trained on such data are more likely to produce outputs that reflect these biases."(p. 78)

Entity— Who or what caused the harm

Human

Due to a decision or action made by humans

AI system

Due to a decision or action made by an AI system

Other

Due to some other reason or is ambiguous

Intent— Whether the harm was intentional or accidental

Intentional

Due to an expected outcome from pursuing a goal

Unintentional

Due to an unexpected outcome from pursuing a goal

Other

Without clearly specifying the intentionality

Timing— Whether the risk is pre- or post-deployment

Pre-deployment

Occurring before the AI is deployed

Post-deployment

Occurring after the AI model has been trained and deployed

Other

Without a clearly specified time of occurrence

Supporting Evidence (1)

1.
"Biases in training data are likely to “disproportionately align with existing regimes of power.”352 For example, prior to the #MeToo movement, the internet was influenced by male-dominated institutions and media that downplayed gender-based violence. Algorithms and content moderation amplified voices aligned with these power structures, giving minimal space to allegations of sexual misconduct."(p. 78)

Other risks from G'sell (2024) (33)

Technical and operational risks

7.3 Lack of capability or robustness
AI systemUnintentionalOther

Technical and operational risks > Technical vulnerabilities (Robustness - unexpected behaviour)

7.3 Lack of capability or robustness
AI systemOtherPost-deployment

Technical and operational risks > Technical vulnerabilities (Robustness - vulnerability to jailbreaking

2.2 AI system security vulnerabilities and attacks
HumanIntentionalPost-deployment

Technical and operational risks > Technical vulnerabilities (The risk of misalignment)

7.1 AI pursuing its own goals in conflict with human goals or values
AI systemOtherPost-deployment

Technical and operational risks > Factually incorrect content (inaccuracies and fabricated sources)

3.1 False or misleading information
AI systemUnintentionalPost-deployment

Technical and operational risks > Opacity (the black box problem)

7.4 Lack of transparency or interpretability
OtherUnintentionalOther
View all 33 risks from this paper →