All micro-internships
Micro-internship brief ≈7h in one sittinglow-code

Analyse a group chat — with consent and anonymisation

A WhatsApp group export is a real behavioural dataset sitting in your pocket. Get the group’s consent, anonymise it properly, then use AI to analyse how the group actually behaves — the privacy part is graded as seriously as the analysis.

Data AnalysisSpreadsheet AnalysisData VisualisationAnalytical Reasoning

Posted by The AI Internships

The work brief

  1. 01Ask a group you’re in for permission (a simple message and yes-replies count — screenshot it). Explain what you’ll analyse and that names will be anonymised.
  2. 02Export the chat (WhatsApp → Export without media) and anonymise: replace every name with Member1, Member2… before any analysis.
  3. 03Analyse with AI (ChatGPT data analysis or Gemini) — feed it only the anonymised export, never real names: messages by hour and day, most active members, response gaps. Verify its counts against the raw file before you believe them.
  4. 04Report what the group would find funny and true — and one thing the exercise taught you about data privacy.

What you’ll produce

4 deliverables

Submission standard

Submit the consent evidence, your anonymised summary stats as CSV, findings, and a chart. Any real name anywhere in the submission is an automatic fail. Include your AI workflow — a real name pasted into an AI prompt counts as a leak. Attach a screenshot of the consent exchange with all names and numbers blurred — an unproven or de-anonymising submission fails.

  • How you got consent (describe the message + who agreed) and your anonymisation steps

    Describe the consent ask and who agreed, and your anonymisation steps — and back it with the consent screenshot (identities blurred).

    Written responseRequired

  • Anonymised summary stats as CSV (by hour/day/member — Member1, Member2…)

    Anonymised summary stats (by hour/day/member — Member1, Member2…) that are derivable from the raw export; counts must reconcile with your findings.

    CSV dataRequired

  • The findings: activity patterns, response behaviour, 3 things the group doesn’t know about itself

    Written responseRequired

  • A chart of the group’s activity pattern

    Image uploadRequired

You’ll complete these inside your private workspace.

What you must submit as proof

This brief requires evidence an AI can’t fabricate.

  • Screenshot of the consent exchange (names/numbers blurred) — the ask and the yes-replies

    Blur or crop every real name and number; we need to see genuine consent, not identities. Any visible real name is an automatic fail.

    Image uploadRequired

Submissions without this evidence cannot be submitted.

Protect other people in your proof. Blur faces, names, phone numbers and email addresses before you upload, and refer to anyone you worked with by role or number ("Listener 1", "the stall owner"). Your proof is only ever used to check your work — it is never published, never appears on your certificate, and is never shown in your public portfolio.

How your work is evaluated

The passing benchmark is 70/100.

Consent & anonymisation

29%

Consent was really obtained and no identifying detail survives anywhere.

Analysis depth

29%

Multiple real dimensions analysed (time, members, response gaps).

Insight

14%

Findings are specific enough that the group would recognise itself.

Proven, clean consent

29%

A screenshot must show real consent was asked and given, with every identifier removed, and the stats must be consistent with a real anonymised export. Consent claimed only in prose, or any leaked real name, is a fail.

How we grade your AI usage

30% of your score

Using AI is the point — it’s the skill this certificate proves. You’ll answer three short questions about how you used it: what you asked, what was wrong with its first answer, and what you changed. Specific, honest answers score high. “I pasted the brief and submitted the answer” scores near zero.