Concerns over undisclosed use of artificial intelligence in literary works have prompted growing efforts to certify books as authored solely by humans. Various labeling initiatives were highlighted at this year’s Frankfurt Book Fair, the world’s largest gathering of the publishing industry, where debates over the impact of AI on writing were prominent.

The movement toward “human-authored” certifications follows several controversies involving authors suspected of employing AI tools without disclosure. Among the most high-profile cases is that of Haitian-Canadian writer Thelyson Orelien, who faced allegations of using AI in his work, a claim he has denied. The controversy has been described as a significant shock to the French-language book market. Charlotte Allibert, co-founder of the French self-publishing and literary agency Librinova, characterized the incident as a “real earthquake” for the industry.

Librinova operates a labeling program in France under which authors may have their books reviewed and certified as “human creations” for a modest fee. To qualify, works are evaluated through AI detection software, and the authors undergo interviews with auditors who conduct a manual review of the text. Approximately 200 titles have received the certification to date. Allibert emphasized that the initiative aims to provide reassurance to authors by demonstrating a firm commitment to addressing AI’s challenges in literature.

In the United Kingdom, the startup Books By People has launched a certification mark for human-authored books, which it recently registered with the UK Intellectual Property Office. The program’s criteria require that writing must originate from the author, although limited AI assistance is allowed for mechanical tasks such as spelling or grammar correction. Co-founder Esme Dennys described the effort as an attempt to establish a “comprehensive, transparent, reliable framework” for publishers and authors amid rising concerns about undisclosed AI use. The scheme, which began with five independent publishers, operates on a fee basis and plans to expand its membership in the near future.

Across the Atlantic, the Authors Guild in the United States has introduced its own human-authored certification mark. Unlike the European programs, it functions as a self-certification system in which authors sign a license agreement verifying their work’s originality without undeclared AI involvement. To date, about 6,500 titles have been awarded this stamp.

As artificial intelligence tools become increasingly accessible, the publishing industry is intensifying efforts to maintain transparency and uphold trust between authors, publishers, and readers. The various human-authored labeling schemes reflect broader attempts to navigate the ethical and practical complexities raised by AI integration in creative writing.