Author Accused of Using AI Because of the Font She Chose

Author T.M. Kirk says she was accused of using AI in an online post because of the font she chose. Kirk said the post carried no AI flag, yet commenters blamed her for making it look machine-made. The episode is small, but it points to a wider problem: the clues people use to spot AI-made work are getting weaker, and honest creators are paying the price.

Tools such as ChatGPT and Claude have made it quicker and cheaper to create graphics. Their spread has prompted calls to boycott businesses that use what critics call “AI slop” in advertising. As readers hunt for signs of machine-made work, familiar design choices can draw the same suspicion.

What makes a font look AI-made

When an image generator draws lettering as part of a picture, it produces the letters as pixels, resembling common styles from its training data. Small inconsistencies in spacing or letter shapes can give the text an uneven look.

Some AI-generated lettering resembles Inter, a clean sans-serif face used widely on screens. Slanted brush lettering in dramatic ads has drawn comparisons to Road Rage. Designer Ramin Nasibov highlighted the brush style in a November 2 post. Human designers can choose the same style, so the lettering alone tells readers little about how an image was made.

Creators are changing how they work

The scrutiny has moved beyond images to em dashes and words such as “delve”, which readers treat as signs of chatbot prose. Some writers use more casual language, leave in typos or avoid certain words, just to sound human. Artists share sketches, saved layers and time-lapse videos to document their work, though AI can now imitate even a time-lapse.

Some designers avoid typefaces that readers associate with AI. As generators get better at clean lettering, those visual clues will become less useful, but the stigma around particular fonts may last longer. The cost falls hardest on small creators: a large publisher can ride out a false claim, while an independent author selling books through Instagram cannot.

Accused of using AI: false claims are now a pattern

Kirk is not the first author to face this. Zimbabwean writer Simbarashe Kundizeza won the Caine Prize on October 6 for his story “The Blue Fighter”, only for detection software to judge almost all of it machine-generated. After a month-long probe, organisers ruled on November 11 that it was human-written. Kundizeza had said from the start: “For the avoidance of doubt, I wrote ‘The Blue Fighter.’ I did not use generative AI to write it. It is entirely my own work.”

Bloomberg has reported the same pressure from the other side: writers are removing em dashes and deliberately adding typos so AI detectors will not accuse them, and authors have lost million-dollar book deals over AI accusations. Students face the same pressure in a different form: the Higher Education Commission has made a skill test mandatory for IT students from 2027, so young writers and coders are judged on provable ability rather than style alone.

What readers should check instead

Style tells are poor evidence. A font, a dash or a tidy sentence proves nothing about how a piece was made. Better checks exist: dated drafts, notes, earlier work in the same voice, and the author’s own account of the process. Google’s Digital Sahafat programme has trained 4,000 Pakistani journalists in AI verification tools, a reminder that checking claims takes method, not instinct. A recent study found AI chatbots can send shoppers to fake stores, and the PTA counted 326 cyber attacks in a year.

Kirk’s post carried no AI flag, and the commenters had no proof. Until detection catches up with generation, that gap will keep producing cases like hers.

Frequently asked questions

Who is T.M. Kirk?

T.M. Kirk is an independent author who promotes her books online. Commenters accused her of using AI because of her font choice.

Can a font prove that AI made an image?

No. Generators imitate styles from their training data, so their lettering can resemble real typefaces such as Inter. A human designer can pick the same styles.

Why are writers changing how they write?

Many writers avoid em dashes, certain words and tidy phrasing because readers treat them as signs of AI writing. Some leave in typos to sound human.

Has anyone been cleared after a false AI accusation?

Yes. The Caine Prize probed Zimbabwean author Simbarashe Kundizeza for a month and ruled on November 11, 2026, that “The Blue Fighter” was human-written.

How can readers check AI claims fairly?

Look for dated drafts, notes and earlier work in the same voice. Style clues alone, such as a font or a dash, are not reliable evidence.