One parent. One investigation. Every claim sourced.

Who I am

My name is Ben Fitzpatrick. I am a parent at a NSW public school. I work in digital marketing and technology. In early 2026, I started asking questions about my school's Facebook Page after reading that Meta had told the Australian Senate it scrapes every public post to train AI.

What I found was not about my school. It was about every school in NSW with a public Facebook Page. The same consent framework. The same policy. The same gap between what parents agreed to and what actually happens to publicly accessible photos in 2026.

I built this site to share the evidence. I do not speak on behalf of any school, P&C, or parent group. I speak as one parent who looked at the data and decided other parents should see it too.

Ben Fitzpatrick on LinkedIn

How this site was built

This site was built collaboratively with Claude Code, an AI coding tool made by Anthropic. The research methodology, data analysis, writing, and site design were produced through a working partnership between me and Claude Code. I directed the investigation, made every editorial decision, and verified every claim. Claude Code helped me move faster and further than I could have alone.

I am transparent about this because it matters. This is a site about technology and its consequences. The tools I used to build it are part of the story.

I am not anti-AI. I am anti-exploitation.

I believe AI is one of the most important technologies of our generation. The tools being built right now, including the one I used to build this site, are genuinely transformative. I am all in on what this technology can achieve.

But AI companies and social media companies are not the same thing. AI companies are building new capabilities. Social media companies have spent two decades proving they cannot be trusted with the data people give them. Meta did not ask permission to scrape every Australian's public posts for AI training. It just did it. When asked why Australians do not get an opt-out, Meta's answer was that Australian law does not force one. Clearview AI scraped 50 billion photos from Facebook, was found in breach of the Privacy Act, ordered to delete the data, and the database grew after the order.

The problem documented on this site is not that AI exists. It is that social media companies have made children's photos publicly accessible by default, under consent frameworks that do not mention AI, governed by policies written before any of these risks existed. The technology is not the failure. The governance is.

Principles

• Every claim is sourced. Senate testimony, regulatory findings, published research, company admissions. If I cannot source it, I do not publish it.
• No blame. Schools are following the system as designed. Parents signed forms in good faith. The gap is between the consent framework and the current reality.
• Precision over heat. The facts are alarming enough. Where I am uncertain, I say so. Where data has limitations, I state them.
• I never access children's photos. All research uses metadata, public indexes, and policy documents. No content from any school Facebook Page was viewed, downloaded, or scraped. This is a hard rule and a methodological strength.
• Action over despair. Every problem section leads to something a reader can do. The goal is not to frighten people. It is to inform them and give them a path forward.

Funding and independence

This site has no funding. No sponsorship. No advertising. No commercial relationships. The domain, hosting, and tools are paid for personally. I have no financial interest in any outcome. I am not affiliated with any political party, advocacy group, technology company, or competitor to Facebook.

If this work is useful, I ask only that you share it. If you want to support it, act at your school. That is worth more than money.

Contact

Email: hello@algorithms.technology

I welcome corrections, additional evidence, and collaboration from researchers, journalists, and policy makers. If I have something wrong, I want to know.

For media enquiries, please email with "Media" in the subject line. I can provide the full dataset, methodology details, and quotes on the record.

Corrections

If any factual claim on this site is incorrect, incomplete, or misleading, please contact me. I will investigate and correct it promptly. All corrections will be noted with the date of change. Getting the facts right is more important than being right.