Ethics of Behavioural Influence and Prediction
PI: Dr Thomas Douglas
The Programme on the Ethics of Behavioural Influence and Prediction (EBIP) is based in the Oxford Uehiro Centre for Practical Ethics. It investigates the moral permissibility and desirability of (i) predicting how people will behave, for example, on the basis of data about their past behaviour, demographic characteristics, and neurobiology, and (ii) influencing how people will behave, for example, through the use of nudges, incentives, psychological interventions, and psychopharmaceuticals.
Questions of interest include:
- What are the advantages and disadvantages of algorithmic forms of behaviour prediction (as compared to discretionary approaches based on clinical/judicial judgement)?
- What would an ideally fair behaviour prediction algorithm look like?
- What sorts of data may permissibly be used as an input in to behaviour prediction algorithms? Demographic variables? Past behaviour? ‘Big data’? Biological factors?
- Do machine learning approaches to behaviour prediction raise new ethical issues?
- What are the ethical similarities and differences between biological and environmental forms of behavioural influence?
- What are the ethically salient categories of behavioural influence? How useful, for ethical discussion, are the categories of nudging and manipulation?
- Are there always reasons to prefer rationality-engaging over rationality-bypassing forms of behavioural influence?
- Is there a right against nonconsensual behavioural influence (of certain kinds)?
Please see below for information regarding our funded research projects
Protecting Minds
Protecting Minds: The Right to Mental Integrity and the Ethics of Arational Influence
Unlike most traditional forms of behavioural influence, such as rational persuasion, incentivisation and coercion, many novel forms of behavioural influence operate at a subrational level, bypassing the targeted individual’s capacity to respond to reasons. Examples include bottomless newsfeeds, randomised rewards, and other ‘persuasive’ technologies employed by online platforms and computer game designers. They also include biological interventions, such as the use of drugs, nutritional supplements or non-invasive brain stimulation to facilitate criminal rehabilitation. The ethical acceptability of such arational influence depends crucially on whether we possess a moral right to mental integrity, and, if so, what kinds of mental interference it rules out. Unfortunately, these questions are yet to be addressed.
Though the right to bodily integrity is well-established, the possibility of a right to mental integrity has attracted little philosophical scrutiny. The purposes of this project, funded by a European Research Council Consolidator Award, are to (1) determine whether and how a moral right to mental integrity can be established; (2) develop an account of its scope, weight, and robustness, and (3) determine what forms of arational influence infringe it, and whether and when these might nevertheless be justified. The analysis will be applied to controversial novel forms of arational influence including persuasive digital technologies, salience-based nudges, treatments for childhood behavioural disorders, and biological interventions in criminal rehabilitation.
Neurocorrectives
Neurointerventions in Crime Prevention: An Ethical Analysis
Interventions that act directly on the brain, or ‘neurointerventions’, are increasingly being used or advocated for crime prevention. For instance, drugs that attenuate sexual desire are sometimes used to prevent recidivism in sex offenders, while drug-based treatments for substance abuse have been used to reduce addiction-related offending. Recent scientific developments suggest that the range of neurointerventions capable of preventing criminal offending may eventually expand to include, for example, drugs capable of reducing aggression or enhancing empathy.
In this Wellcome Trust-funded project, we are investigating ethical questions raised by the use of such interventions to prevent criminal offending, focusing particularly on cases where they are imposed on convicted offenders as part of a criminal sentence or as a condition of parole. On the one hand, there seems to be at least some reason to support the use of neurointerventions in this way, since there is a clear need for new means of preventing crime. Traditional means of crime prevention, such as incarceration, are frequently ineffective and can have serious negative side-effects; neurointervention may increasingly seem, and sometimes be, a more effective and humane alternative.
On the other hand, neurointerventions can be highly intrusive and may threaten fundamental human values, such as bodily integrity and freedom of thought. In addition, humanity has a track record of misguided and unwarrantedly coercive use of psychosurgery and other neurotechnological 'solutions' to criminality.
We are deploying philosophical methods and recent thinking on autonomy, coercion, mental integrity and moral liability to answer two over-arching questions
- When, if ever, may the state force neurointerventions on criminal offenders?
- When, if ever, may the state offer neurointerventions to criminal offenders?
We plan also to examine how our answers to these questions bear on the use of neurointerventions to prevent offending in individuals who have not previously offended, but are thought to be at high risk of doing so.
