A leaked slide deck put Instagram's own teen research on trial

The short answer

In September 2021, reporting on leaked internal slides suggested Meta's own research showed Instagram was harmful to some teenage girls' body image. Meta published its own account of that research on 26 September 2021, and a Senate subcommittee questioned the company's global head of safety about it four days later. The episode shows how differently a small internal survey can read once it becomes public and contested.

What the evidence says

In its post, What Our Research Really Says About Teen Well-Being and Instagram, Meta stated that across 12 well-being topics its internal surveys asked about, teenage girls who said they struggled with 11 of them also said Instagram made that struggle better rather than worse; body image was the exception, where more girls with body image concerns said the app made things worse, though even there a large share reported no change or improvement. Meta said the research did not establish a causal relationship between Instagram use and mental-health outcomes, and disclosed that some of the underlying focus groups involved as few as 40 teenage participants.

On 30 September 2021, the Senate Commerce Committee's consumer protection subcommittee held a hearing, Protecting Kids Online: Facebook, Instagram, and Mental Health Harms, questioning Facebook's Antigone Davis about the company's research into its apps' effects on young users and what actions it had taken in response.

For context

What became public was Meta's own description of its own research, produced under public and congressional pressure rather than released as an independent academic study. A focus group of around 40 teenagers is a useful tool for surfacing themes, but it cannot generalise to all teenage Instagram users the way a large probability survey can, and neither the leaked slides nor Meta's rebuttal have been independently replicated by outside researchers with access to the same raw data.

A practical next step

Rather than treating a single leaked or company-published account as the full picture, you could ask your own teenager directly how a specific app makes them feel about their appearance or their day, since that conversation captures something neither a small internal study nor a public hearing can.

• Does my teenager say a particular app changes how they feel about their appearance, and in which direction?
• What settings, such as time limits or content controls, does that app already offer, and are they turned on?
• Would my teenager be willing to take a short break from one app to see whether their own mood shifts?

This remains a story about disputed evidence and institutional accountability as much as about screens themselves: a company describing its own findings, and a legislature testing that account in public, neither of which substitutes for independent research into how any one child experiences a given app.