"AI writing software" and "traditional writing software" get talked about like two points on the same scale, more or less automated. They're closer to two different categories of tool doing genuinely different jobs, and conflating them makes it harder to pick the right one for how you actually want to write.
What an AI-assisted tool actually does
Tools like Novelcrafter or Sudowrite generate text, whole paragraphs, plot suggestions, dialogue, based on prompts and context you provide, using a language model connected either through the app itself or your own API key. The model has been trained on large volumes of text and produces new text statistically likely to fit what you've asked for. This is genuinely useful for writers who want a collaborator to draft against, brainstorm with, or push past a stuck scene, and it comes with real, ongoing costs (subscription plus, often, separate API usage) and open questions about how the output should be credited or disclosed.
What a traditional writing tool actually does
Tools like Scrivener, Dabble, or Quillen organize, format, and hold what you write, without generating any of the prose itself. Structure (chapters, scenes, outlines), formatting (industry-standard manuscript layout), and organization (characters, world details) are automated; the words themselves are entirely yours, typed by you, start to finish.
Where rules-based analysis sits in between, sort of
A tool like Quillen's Analyze Document feature complicates the simple two-category picture slightly. It automatically flags filter words, passive voice, and other prose habits directly in your manuscript, which sounds AI-adjacent. It isn't: it works by checking your text against a fixed, human-written list of patterns and word categories, not a trained model, doesn't generate or rewrite anything, and doesn't send your text anywhere to be processed. It's automation applied to feedback, not automation applied to writing.
The question worth actually asking yourself
Not "which category is better," but "do I want a collaborator that generates prose, or a tool that organizes and formats the prose I write myself." Some writers genuinely want the former, especially for brainstorming or breaking through a stuck draft. Others want every word to be unambiguously their own, with software handling only the structural and formatting work around it. Neither preference is wrong, and the two kinds of tool aren't really in competition with each other so much as built for different relationships to the actual writing.
Where disclosure policies fit into this
Some publishers and writing contests now require disclosure of AI-generated content in a submission; others don't address it directly yet. Policies here are genuinely still settling industry-wide, so if this matters for a specific submission path, checking that agent, publisher, or contest's current stated policy is worth doing rather than assuming a universal standard exists.
Common questions
Do agents and publishers care if I used AI writing software?
Policies vary and are still evolving industry-wide. Some publishers and contests now require disclosure of AI-generated content; others focus only on the final prose. If this matters for your specific submission path, check the individual agent's, publisher's, or contest's current policy rather than assuming a blanket standard.
Is rules-based prose analysis the same thing as AI editing?
No. Rules-based analysis checks text against a fixed, human-written list of patterns (specific words, sentence structures) and flags matches. It doesn't generate or rewrite text, learn from data, or send anything to an external model, which is a fundamentally different mechanism from an AI editing tool.