PublicLab

• How can we recognise when people are being used as part of a digital experiment?

• What should be the rules to protect people when experimentation occurs digitally instead of in-person?

PublicLab aims to identify the characteristics and thresholds for problematic digital experimentation and to set the agenda for governing it. By doing so, the project seeks to strengthen accountability and foster innovative governance around digital experimentation worldwide.

The PublicLab project aims to better protect people subjected to digital experimentation, bringing it into line with protections in traditional research. The project focuses on ethics in digitised health and social science research, as well as experimentation carried out within policy and corporate environments. It examines when digital experimentation becomes problematic and identifies emerging risks.

PublicLab will also study how international networks of public and private actors collaborate to test policies, products or ideas. It will produce a new framework to improve the governance of digital experiments, particularly in hybrid settings involving multiple types of organizations.

Motivation for PublicLab

Governance gaps and mismatched problem definitions have created a paradox: digital experiments can pose serious risks to people’s rights and welfare while also satisfying research ethics rules, preserving privacy and data security, employing explainable algorithmic processes and being robust and statistically unbiased. 

There is a pressing need to find out why this is so, and to come up with new ways of governing experimentation that take digitisation into account.

Why is this PublicLab relevant?

Today, everyone alive is part of continual remote digital experimentation conducted by business, government and researchers. Experiments are conducted in real time through the digital devices we use, through generative AI, but also through more traditional policy and governmental modes such as policy nudging, trials and experiments in international development settings, and in the humanitarian sphere.

Recognizing when experiments that are invisible and impossible to opt out of or become harmful is a necessary first step to distinguishing problematic experimentation and learning how to regulate it. 

Expected PublicLab outputs

The project will produce:

• A global taxonomy of datafied forms of experimentation (open dataset)
• A typology that will define the new categories of digital experiment
• Two regional workshops
• A series of coauthored papers with visiting fellows from different regions
• A methodological playbook for critical research on infrastructures of experimentation
• A special issue on governance of experimentation
• Working papers providing a governance roadmap
• An open-access edited volume for an interdisciplinary academic audience
• A seminar series open to ISS and researchers globally