What harms can AI cause?
From education and healthcare to finances and employment, AI and data-driven technologies are increasingly used to make and support important decisions about our everyday lives.
AI systems are likely to have benefits, but they can also cause harm to people, society and the environment. This means that there are important ethical and legal issues to consider in relation to their use.
These issues include potentials for causing or increasing:
lack of explainability and transparency – this is when there is not enough clear information explaining how AI systems operate and how some decisions are made
toxic, abusive or harmful outputs
privacy, data protection and surveillance issues
copyright and intellectual property issues
inaccuracies and loss of human judgement
fraud, scams and manipulation
economic inequalities
environmental harms.
Some of these issues relate to the nature of AI systems and how they are developed. For example, sometimes there is a lack of transparency and ‘explainability’ about how datasets and machine-learning algorithms are created.
Other risks relate to how these AI systems are used and what they are used for. This can include examples such as the generation of ‘deepfakes’, and the use of biometric technologies to scan and identify people.