From the Periphery: Typed by Hand ISTE 2026

Let me tell you about the problem with AI detectors.

They are trying to answer the wrong question.

Every AI detector on the market: GPTZero, Turnitin, Copyleaks, all of them, asks some version of the same question: does this finished text look like it was written by a human or a machine? They analyze the output. They look for patterns. They generate a score. And then they tell you, with varying degrees of confidence, whether the thing already submitted to you is probably authentic or probably not.

The problems with this approach are well documented and genuinely serious. False positives that flag non-native English speakers at disproportionate rates. Accuracy claims that are often self-reported and rarely independently verified. Students who have learned to "humanize" AI output specifically to defeat detectors. The fundamental reality that the more sophisticated AI becomes, the less reliably any output-based detector can catch it.

And underneath all of that, a deeper problem that we do not discuss enough: AI detectors are reactive. They look at what already happened. They cannot tell you anything about how the writing came to be, only whether the finished product pattern-matches to their training data on AI versus human text.

Typed by Hand has a completely different focus, and I think it could be the right one.

What is Typed by Hand?

Typed by Hand is not an AI detector.

Typed by Hand is a writing verification platform. Students write directly inside it, and as they write, the platform tracks the process in real time. Keystrokes. Tab switching. Copy and paste events. The actual, moment-by-moment record of how a piece of writing came to exist.

The result is not a score; it is a story. A paragraph-by-paragraph account of what happened while the writing was being created. Did the student type it character by character over an hour, with the natural pauses and corrections and false starts that authentic writing produces? Or did 800 words appear in a single paste event followed by minor edits?

Those two things look completely different in Typed by Hand. And unlike an AI detector score, they do not require interpretation or probabilistic reasoning. You can see what happened. The process is the evidence.

This is the fundamental insight that makes Typed by Hand interesting: you cannot fake a natural writing process. You can rewrite AI text to fool a content analyzer. You can run it through a humanizer to defeat a detector. But faking two hours of organic writing with realistic keystroke speed, natural pauses, genuine corrections, and authentic revision is a different challenge entirely. The process is significantly harder to spoof than the product.

https://www.youtube.com/watch?v=MJXcdUG-hZc

For teachers, this means something significant: you are no longer in the position of accusing a student based on a probabilistic score that a well-coached defense attorney could challenge. You have a record. A timeline. A paragraph-by-paragraph account of how the work came to exist. That is not the same kind of evidence. That is better evidence.

LinkedIn Piqued my Interest

Typed by Hand was not a surprise discovery in the same way that DeskPad was. I had already seen them on LinkedIn before ISTE 2026 and made a mental note to find their booth.

That is its own kind of endorsement, honestly. A tool that makes you seek it out before you even arrive at the conference is a tool that has communicated its value proposition clearly enough that you remember it in a crowded field.

What the booth visit confirmed was what I had suspected from the LinkedIn post: this is a team that understands the actual problem with academic integrity in an AI era. The problem is not that students are using AI. The problem is that our current tools for responding to it are either blunt instruments with serious accuracy and equity problems or Socratic approaches like DeskPad that work beautifully but require students to opt in to the right environment.

Typed by Hand is a different kind of solution. It creates a writing environment where authentic writing is verified by the process rather than policed by the output.

What This Is Not

Typed by Hand is a monitoring tool. Students are writing inside a platform that is tracking their behavior. That raises legitimate questions about privacy, about student trust, and about whether surveillance is the right frame for academic integrity at all.

These are not questions I am going to answer for you. They are questions your school, your district, and your students deserve to be part of answering. What I will say is that Typed by Hand is transparent about what it tracks, and that transparency is more than can be said for some of the tools already deployed in schools. However, I do know this is the type of ecosystem that teachers are looking for before they revert to pencil and paper only. I think it gives those types of teachers a more accurate and equitable tool to ensure that they are receiving student-created, authentic work.

The argument for it is not that monitoring is good. The argument is that if schools are going to address academic integrity in an AI era, and they have to, then verifying the process of writing is significantly more accurate, more fair, and more defensible than analyzing the finished product and guessing.

That is the argument that stopped me at their booth. And I still think it holds.

Try It Yourself

🔗 https://www.typedbyhand.com