Taca

A short study exploring how we respond to frustration online.
Takes about 3 minutes.

Before you begin — what this tool does with your data
  • You'll answer two brief questions and have a short conversation with an AI.
  • Your scale responses (1–5 scores) and the number of messages you exchange with the AI are recorded anonymously in aggregate — with no name, email, or identifier linking them to you.
  • Your chat messages are sent to the Anthropic API for processing and are not stored by Taca after your session ends. Anthropic does not use API conversations to train its models.
  • For security and abuse prevention, our server briefly processes each request's IP address (via FoxxeLabs Sentinel). IP addresses are not stored alongside your responses and are not used to identify you.
  • A theme preference may be saved in your browser's local storage. No other cookies are set.
  • Legal basis: academic research (ATU Donegal, MSc Cyberpsychology, 2026). Data controller: FoxxeLabs Limited, Letterkenny, Co. Donegal.
Privacy policy  ·  About this research  ·  FoxxeLabs

How would you respond?

We're going to show you a common online situation. There's no right or wrong response — we're interested in your honest reaction.

You post a comment sharing your opinion on a topic you care about. Someone you don't know replies with a dismissive, mocking response that ignores your actual point and makes fun of your view in front of others.

Tell us what's been frustrating you

Before we show you another situation, we'd like to hear about something that has been frustrating you online lately — anything at all. This is just between you and the AI. Your conversation is not stored.

One more situation

Same question as before — what would you most likely do?

You're in an online group for a hobby you enjoy. Someone posts confidently incorrect information. When you politely correct them with a source, they respond aggressively, accusing you of being condescending and trying to start an argument.
🙏

Thank you for taking part

Your responses have been recorded. Here's a little more about what this study is exploring.

What is Taca investigating?

Neuroscientist Lisa Feldman Barrett (2020) argues that the nervous system is partly regulated by the people around us — their presence, responsiveness, and predictability directly affect how we feel and how we're likely to act. Online environments strip that away. Anonymity removes the social cost of hostile expression, and without other people present to co-regulate us, frustration travels more directly toward aggression.

In research data from 173 Irish university students (McCaffrey, 2026), participants who used AI to vent frustration showed a significantly lower likelihood of hostile online responding — even after controlling for other predictors. Crucially, the effect was specific to venting. Other forms of AI use showed no relationship at all.

Taca is exploring whether a responsive AI can partially restore the regulatory function that online anonymity removes — functioning not just as a tool, but as a new kind of social participant. Your responses contribute to the next phase of that research.

Read the research →
Taca is a Phase 2 public demonstrator. Response data is anonymised and retained only in aggregate.
Built by FoxxeLabs · UCA dissertation research, ATU Donegal, 2026 · Literature review · Privacy policy