This session will conduct an experiment on the state of privacy and AI transparency in Internet. We will introduce a set of visual signs designed to replace current consent and privacy online forms (see https://www.linkedin.com/pulse/how-inform-users-intuitively-non-intrusively-rights-garcia-gasulla/).
The experiment will engage participants with current consent and privacy forms of popular webpages to assess their accessibility and actual degree of transparency. This will be performed by participants, as they assign the proposed visual signs to popular websites. Discussions will be encouraged, to identify general tips (e.g., keywords to look for, how much effort it implies, form traps and how to avoid them, etc.).
If the results are of interest, a scientific paper will be produced. All participants will be acknowledged in the paper, and participants who can actively contribute to the manuscript will be invited as co-authors.
If we manage to build a community around the idea, we could empower citizens by providing the appropriate signs for many different AI services. This would motivate the owners of such services to be more transparent and open in the near future.
What is the goal and/or outcome of your session?:First, creating awareness on the AI community regarding transparency of AI terms and conditions (or the lack of).
Second, creating a working group around the visual signs which could guide the general public in their interactions with AI.
A potential outcome may be a collaborative paper, which we envision as a practical exercise of our previous theoretical work.
Depending on the number of participants, we may be able to guide them more thoroughly or less. If there are many participants, one step-by-step example will be provided, and then individual questions will be solved. If there are less participants, we may be able to work hand-in-hand with them, so that each can cover a larger number of target sites and services.
I'm a PostDoc researcher specialized on AI. I teach at UPC. I'm interested in many aspects of AI, including trustworthy AI.
PhD in AI, researcher at Barcelona Supercomputing Center, co-director of AI4EU Observatory for Society and AI. 10+ years involved in EU funded projects related to AI and eHealth.