---
title: Publicly funded AI should strengthen relatedness, competence and autonomy
author: Martin Ebner
date: 2026-11-04
url: https://postgenerative.at/indications/2026-11-04/publicly-funded-ai-should-strengthen-relatedness-competence
lang: en
type: Indication
summary: "A study calls for well-being as an EU AI policy goal: subsidized companies should strengthen relatedness, competence and autonomy and substantiate their societal benefits."
tags:
  - Politics
  - Research
sources:
  - title: "Weizenbaum Institute: Public money for AI – but what about happiness?"
    url: https://www.weizenbaum-institut.de/news/detail/neue-studie-optimising-happiness/
  - title: "Wachter, Mittelstadt, Onitiu and Weinstein: Optimizing for happiness in EU technology policy to repay public subsidies for AI"
    url: https://doi.org/10.1098/rsos.261323
license: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)
alternate:
  de: https://postgenerativ.at/indizien/2026-11-04/oeffentlich-gefoerderte-ki-soll-verbundenheit-kompetenz-und
---

# Publicly funded AI should strengthen relatedness, competence and autonomy

Public support for AI should come with demonstrable societal benefits. The Weizenbaum Institute presents this proposal by Sandra Wachter, Brent Mittelstadt, Daria Onitiu and Netta Weinstein. [1] Their paper draws on psychological research and an analysis of AI policy; it is not a new experiment testing the effects of individual AI products. [2]

Drawing on self-determination theory, the authors identify three basic psychological needs as foundations of happiness and well-being:

- **Relatedness:** People need meaningful relationships, care and belonging. Connection to nature also matters. AI services should support these connections; engagement-driven chatbot design and loss of access to nature through data centres can work against them.
- **Competence:** People seek to learn, master challenges and experience themselves as effective. AI should therefore support skills and meaningful activity. Persistently outsourcing thinking and learning tasks risks deskilling.
- **Autonomy:** This means acting willingly and in accordance with one’s values. Receiving support from others is compatible with it. Manipulative personalization, excessively agreeable chatbots and pressure can undermine self-determination. [2]

Companies should use these needs as design goals without paternalistically forcing users into supposedly happy behaviour. For public subsidies, the authors propose detailed plans specifying expected benefits, justified design choices, use cases, metrics and scientific evidence. Companies should be accountable for delivering on their predictions. Binding standards for implementation, measurement and enforcement still need to be developed. [2]

The societal significance extends beyond personal feelings: the research reviewed connects well-being with learning, creativity, helping others and cooperation. The three needs create conditions enabling people to contribute to their communities. [2]

Read as an indication, this changes the expectations placed on AI subsidies: success should also be judged by whether people can shape their lives with connection, competence and self-determination.