27 canonical risk pages

Society

Collective and system-level impacts such as bias, inequality, information harm, and social damage.

Autonomous Military AI

Development of lethal autonomous weapons systems (LAWS) capable of selecting and attacking targets without significant human intervention, eliminating human control over life-or-death decisions.

MdSeverity 7/10

AI Medical Error

Erroneous diagnoses, inadequate treatment recommendations, or biases in medical AI systems due to unrepresentative datasets or model limitations.

JuSeverity 7/10

Biased Sentencing

Perpetuation and amplification of racial and socioeconomic biases in recidivism prediction systems and automated judicial decision-making (e.g., COMPAS).

NuSeverity 7/10

DeepNude

Non-consensual generation of synthetic nudity images or deepfake pornography of real individuals, constituting image-based sexual abuse.

EbSeverity 7/10

Filter Bubbles

Recommendation algorithms that selectively reinforce the user's pre-existing beliefs, creating echo chambers that amplify polarization and ideological isolation.

HzSeverity 7/10

Hate Speech

Automated generation or amplification of toxic content, targeted harassment, and hate speech via AI systems, facilitating harassment campaigns at scale.

PoSeverity 7/10

Political Polarization

Amplification of political division via extremely personalized microtargeting campaigns generated by AI exploiting individual cognitive biases.

RcSeverity 7/10

Racial Bias

Unequal and discriminatory performance of facial recognition systems and other algorithms on people with darker skin tones, perpetuating systemic racism.

SbSeverity 7/10

Social Bias

Reproduction and amplification of systematic social prejudices present in training data, manifesting as discrimination based on race, gender, age, or other protected characteristics.

TrSeverity 7/10

Truth Erosion

Epistemic collapse caused by the proliferation of synthetic content indistinguishable from authentic content, making reality verification impossible at mass scale.

EdSeverity 6/10

Academic Fraud

Widespread use of generative AI by students to complete academic assignments without developing critical thinking, writing, or problem-solving skills.

EqSeverity 6/10

Access Inequality

Widening of the digital divide due to unequal access to advanced AI technologies, concentrating benefits in privileged populations and excluding disadvantaged communities.

AgSeverity 6/10

Age Bias

Age discrimination in automated hiring systems, credit scoring, and other contexts that unfairly penalize older individuals.

CbSeverity 6/10

AI Cyberbullying

Persistent and automated harassment against individuals using AI bots that operate relentlessly on social networks and digital platforms.

AlSeverity 6/10

Algocracy

Governance via algorithmic systems that make political and administrative decisions without consideration of human context, ethical values, or empathy capacity.

BuSeverity 6/10

Automated Bureaucracy

Automated bureaucratic systems making opaque decisions without effective human appeal mechanisms, creating Kafkaesque mazes of irreversible algorithmic decisions.

CkSeverity 6/10

Clickwork Exploitation

Extreme precariousness of data labeling and annotation work through microtask platforms paying minimal compensation for intense cognitive labor.

CuSeverity 6/10

Cultural Homogenization

Cultural domination of models trained primarily on English and Western content, eroding cultural diversity and marginalizing non-Western perspectives.

GdSeverity 6/10

Gender Bias

Systematic gender stereotypes encoded in AI models that incorrectly associate genders with professional roles, perpetuating discrimination.

GrSeverity 6/10

Group Fairness

Disparity in positive or negative outcome rates between defined demographic groups, constituting systemic discrimination.

FaSeverity 6/10

Individual Fairness

Inconsistent algorithmic treatment of individuals who are similar in relevant aspects, violating principles of individual equity.

LgSeverity 6/10

Linguistic Extinction

Systematic exclusion of languages with scarce digital resources from AI benefits, accelerating the loss of linguistic diversity and extinction of minority languages.

AuSeverity 6/10

Loss of Autonomy

Erosion of human capacity to make informed decisions by delegating excessively to opaque algorithmic systems without understanding their functioning.

NaSeverity 6/10

National Bias

Negative stereotypes and biased representations of specific nationalities and countries, typically reflecting dominant Western perspectives in training data.

RlSeverity 6/10

Religious Bias

Stereotypical associations between specific religions and negative characteristics such as violence or extremism, reflecting prejudices present in training data.

AcSeverity 6/10

Unequal Access

Concentration of access to advanced AI technologies in economically privileged populations, exacerbating existing inequalities.

LtSeverity 5/10

AI Literacy

Widespread lack of public understanding regarding real capabilities, limitations, and risks of AI systems, facilitating disinformation and inappropriate use.