Glossary
AI text detector
An AI text detector is software that estimates whether a passage was written by a language model, usually from the words alone. Detectors make mistakes in both directions, so a score is weak evidence about how a text was made.
Updated October 17, 20264 cited facts, each traced to its source
Also called AI detector, AI writing detector, AI content detector.
An AI text detector is software that estimates whether a passage was written by a language model. It returns a score or a label, such as "likely AI-written." Most detectors judge the words alone, so they cannot see whether you typed, dictated or pasted the text. Schools, publishers and employers use them under names such as AI detector, AI writing detector and AI content detector.
Detectors score the words, not the process
Most detectors are classifiers: a model trained on samples of human writing and machine writing, then used to score new text. Many measure how predictable each word is to a language model, and how much sentence length and structure vary. Machine text often scores as more predictable and more even. Plain, formal or carefully edited human writing can also score as predictable, and that is the root of the problem.
Early detectors were withdrawn for low accuracy
OpenAI withdrew its AI text classifier on July 20, 2023, because of its low rate of accuracy 1. On OpenAI's own challenge set of English texts, the tool correctly identified 26 percent of AI-written text as likely AI-written 2. It missed most of the machine text it was built to find.
Detectors also flag human writing
Errors ran in the other direction too. A 2023 study of several widely used GPT detectors found that they consistently misclassified non-native English writing samples as AI-generated 3. Turnitin, a vendor of plagiarism and writing checks, says that in documents where its AI writing detector finds less than 20 percent AI writing, there is a higher incidence of false positives 4.
A concrete example: a dictated paragraph
Suppose you dictate a short paragraph into a dictation app, then paste it into a detector. The detector sees only the words. If the app left your wording alone, the sentences are still yours, but a detector can still score plain, evenly built sentences as machine-like. If the app used an AI rewrite that changed sentence shape and word choice, the score can shift further. Neither outcome tells the detector how the text was made.
Why it matters for dictation and typing
- Verbatim dictation keeps your words, fillers and all. AI cleanup rewrites them, and that rewriting is the kind of change a detector may react to. Our guide to speech to text without AI explains how to turn the rewrite off.
- A score is not proof. It cannot show who wrote a text or which tool was used. For a teacher or employer, a flag is a reason to ask questions, not a verdict.
- Drafts are evidence. Dictation drafts and document version history show how a text developed, which is more useful than a score. Our question on whether AI detectors can flag dictated text covers this in more detail.
Related terms
- Generative AI: the models that produce the text detectors try to find.
- Large language model (LLM): the usual source of the machine text these tools look for.
- Generative error correction (GER): a language-model pass over speech recognition output, the kind of rewrite that can change dictated wording.
Guides that cover it
Evidence
Every number on this page, with where it comes from. Open a fact to see the exact words in its source.
- 1.
- 2.
- 3.
- 4.
Related terms
- Speech to Text AI: Dictation, Assistants and Voice AgentsThe topic this term belongs to
- Generative AIGenerative AI is the class of AI models that create new content, such as text, audio or images, from patterns in their training data. The NIST profile defines it as models that emulate the structure and characteristics of input data to generate derived synthetic content.
- Generative error correction (GER)Generative error correction (GER) passes a speech recognizer's output, often its N-best list of alternative transcripts, to a large language model, which rewrites the transcript to fix errors. It can recover words the recognizer never offered, but it can also change words that were already right.
- Intent recognitionIntent recognition is the step that decides what a spoken request is asking for, such as setting a timer or sending a message, and which details the action needs. Voice assistants and voice agents rely on it before they act.
- Large language model (LLM)A large language model (LLM) is a language model with a very large number of parameters, trained on huge amounts of text to predict and generate language. Google's glossary defines the minimum as a language model having a very high number of parameters.
- Machine learningMachine learning is a method in which a program learns its behavior from example data rather than from rules written by hand. Google's glossary defines it as a program or system that trains a model from input data.
- Prompt (AI)In AI, a prompt is the input you give a model to get a response, such as a question, an instruction or pasted text. Google's glossary defines it as any text entered as input to a large language model to condition the model to behave in a certain way.
- Speech language modelA speech language model (speech LLM) is a large language model that accepts audio as input, so it can process speech directly instead of working only on a text transcript. The 2023 SALMONN model is one example that lets an LLM process general audio inputs.
- Speech-to-speech modelA speech-to-speech model takes audio in and produces audio out with one model, without writing the conversation out as text in between. Most voice systems instead chain speech recognition, a language model and speech synthesis.
- Text to speech (TTS)Text to speech (TTS), also called speech synthesis, is software that converts written text into spoken audio. Modern systems use neural networks, and they can run on the device or in the cloud.
- Turn detectionTurn detection is the step that decides when a speaker has finished a turn, so a voice system knows when to respond. A pause can be a hesitation rather than an ending, so the decision trades speed against cutting the speaker off.
- Voice agentA voice agent is a voice system that listens, reasons with a language model, uses tools, and speaks back across several turns to finish a task, such as checking an order and rescheduling it. Benchmarks show that agents still fail many tasks on realistic audio.
- Voice assistantA voice assistant is software that listens for a spoken request, works out what you want, and then acts or answers, such as setting a timer or turning off a light. Dictation is different: it turns your speech into text for you to edit.