Ethical and social risks of harm from Language Models

Ethical and social risks of harm from Language Models

Laura Weidinger. John Mellor. Maribeth Rauh. Conor Griffin. Jonathan Uesato. Po-Sen Huang. Myra Cheng. Mia Glaese. Borja Balle. Atoosa Kasirzadeh. Zac Kenton. Sasha Brown. Will Hawkins. Tom Stepleton. Courtney Biles. Abeba Birhane. Julia Haas. Laura Rimell. Lisa Anne Hendricks. William Isaac. Sean Legassick. Geoffrey Irving. Iason Gabriel. 2021. (View Paper → )

This paper aims to help structure the risk landscape associated with large-scale Language Models (LMs). In order to foster advances in responsible innovation, an in-depth understanding of the potential risks posed by these models is needed. A wide range of established and anticipated risks are analysed in detail, drawing on multidisciplinary literature from computer science, linguistics, and social sciences.

It’s great to a have a clear taxonomy and classification of potential harms from language models.

• Discrimination, Exclusion and Toxicity • Social stereotypes and unfair discrimination • Exclusionary norms • Toxic language • Lower performance for some languages and social groups

• Information Hazards • Compromising privacy by leaking private information • Compromising privacy by correctly inferring private information • Risks from leaking or correctly inferring sensitive information

• Misinformation Harms • Disseminating false or misleading information • Causing material harm by disseminating false or poor information e.g. in medicine or law • Leading users to perform unethical or illegal actions

• Malicious Uses • Making disinformation cheaper and more effective • Facilitating fraud, scams and more targeted manipulation • Assisting code generation for cyber attacks, weapons, or malicious use • Illegitimate surveillance and censorship

• Human-Computer Interaction Harms • Anthropomorphising systems can lead to overreliance or unsafe use • Creating avenues for exploiting user trust, nudging or manipulation • Promoting harmful stereotypes by implying gender or ethnic identity

• Automation, access, and environmental harms • Environmental harms from operating LMs • Increasing inequality and negative effects on job quality • Undermining creative economies • Disparate access to benefits due to hardware, software, skill constraints