
AI has quietly entered the writing life. What began as a curiosity — chatbots producing oddly competent sonnets — has matured into something more practical and more interesting. A 2025 NBER study examining real-world ChatGPT usage found that writing accounted for roughly 40% of work-related conversations, making it the platform's largest professional use category. Much of that activity involved editing, critiquing, translation, and summarisation rather than generating entirely new text.
For poets and creative writers, this distinction matters deeply. AI is not replacing the blank page with machine-made verse. It is showing up at specific stages of the creative process — research, structural experimentation, revision, submission preparation — where it can lighten mechanical labour without touching the voice that makes a poem yours. Some literary platforms and publications are going further by weaving generative AI into their editorial systems, sometimes enlisting remote AI developers to connect language models with publishing tools, contributor databases, and review workflows in ways that keep the human editor firmly in charge.
Before the First Line: Research and Inspiration
Research often comes before the poem itself. A poet may need to understand a historical figure more clearly or get a better sense of a place they have never visited. AI can help with that first stage of research. It can explain unfamiliar references and show where different sources tell the story in different ways.
The useful part is what happens next. A detail can suggest an image. A contradiction can create tension. A phrase or fact can become the point from which the poem begins to grow.
That changes the writer's role at the start of the process. Instead of hours spent organising notes, you can move more quickly toward the real question: What does this material mean to me, and what poem lives inside it?
The caution is straightforward. Language models do not distinguish a reliable source from a plausible-sounding one. Verification and interpretation remain yours.Used this way, AI works best as a support tool rather than a substitute for creative judgment.
Harvard Business Review has described generative AI as a technology capable of augmenting human creativity rather than automating it. For poets, that makes AI especially useful as an exploratory companion — surfacing questions you hadn't considered, offering structural alternatives, or mapping the territory of a theme before you commit to a path through it.
Revision Is Becoming More Granular
AI has arguably had a stronger effect on editing than on drafting, and poets stand to gain the most here.
You can now test multiple versions of a stanza without manually rewriting each one. You can compress, expand, shift register, identify unintentional repetition, or compare alternative line breaks in minutes rather than hours.
Common revision uses for poets
|
Task
|
What AI can do
|
What stays with you
|
|
Line-level editing
|
Flag repetition, weak verbs, inconsistent tense
|
Deciding which repetition is intentional
|
|
Structure
|
Suggest reordering stanzas, test alternate forms
|
Choosing the shape that serves the poem
|
|
Tone checking
|
Identify tonal shifts or register breaks
|
Determining whether the shift is a flaw or a feature
|
|
Compression
|
Condense wordy passages
|
Knowing which words earn their place
|
|
Submission prep
|
Format for different journal guidelines
|
Selecting where the poem belongs
|
Research discussed by Harvard Business Review found substantial time savings in professional writing tasks when workers had access to generative AI, while output quality also improved.
The risk for poets is specific: language models normalise text. Instructed to make writing "clearer" or "stronger," they will sand away the unusual phrasing, the fractured syntax, the deliberate roughness that gives a poem its identity. AI suggestions are options, never corrections. A technically smoother line is not always a better line.
Publishing and Submitting in the AI Era
The impact of AI extends beyond the writing itself.
Poets juggling submissions can use AI to:
- Draft and customise cover letters for different journals
- Track submission guidelines and formatting requirements
- Generate poet bios at varying lengths
- Prepare reading descriptions and event summaries
- Adapt a single poem's framing for print, online, and social platforms
For poets publishing chapbooks or collections, AI can assist with sequencing notes, back-cover copy, acknowledgements, and the dozens of small prose tasks that surround a manuscript without being part of it.
The key distinction is between automation and autonomy. Letting AI format your submission packet saves time. Letting it choose which poems to send where, or rewrite your artist statement in a voice that isn't yours, trades something essential for convenience.
As Production Gets Easier, Judgment Matters More
AI reduces the effort of producing text. It does not reduce the cost of publishing work that is derivative, careless, or dishonest.
The easier it becomes to generate passable writing, the more important editorial taste becomes — both your own and the literary community's. Poets need habits for:
- Verifying any factual claim AI helped construct
- Recognising when AI has introduced phrasing or imagery that was never in the source material
- Protecting unpublished work from being fed into systems that may store or learn from it
- Maintaining the line between AI-assisted revision and AI-generated writing
The World Economic Forum's Future of Jobs Report 2025 identifies creative thinking, analytical thinking, and technological literacy among the most important and fastest-growing skill areas — a pattern pointing toward human-AI collaboration rather than replacement.
Knowing how to generate text will become less distinctive. Knowing what deserves to exist as a poem will become more valuable.
The Craft Remains Yours
AI has made parts of the writing process quicker. Poets can test an idea, look into a subject, or work through an early draft in less time. The lasting work still comes from the poet’s eye and ear. It comes through a particular way of noticing an image, hearing the rhythm of a line, and knowing when a stanza needs another revision. A distinct voice grows from those decisions and becomes recognizable on the page.
The strongest creative practice will divide the labour deliberately. AI handles repetitive formatting, supports research, proposes structural alternatives, and accelerates mechanical revision. The poet provides argument, image, music, and meaning. The community — readers, editors, fellow writers — determines whether the result was worth the encounter. The technology changes the mechanics of writing. The craft of making something worth reading remains entirely, stubbornly human.