A new law in Iowa has sparked a controversy over the use of artificial intelligence to filter books in school libraries. The law, known as SF 496, requires schools to remove any books that contain descriptions or visual depictions of a sex act, unless they are deemed age-appropriate by the school board. To comply with this law, some school districts have turned to ChatGPT, a powerful text generation tool developed by OpenAI, to help them identify books that may violate the law.

What is ChatGPT and how does it work?

ChatGPT is a generative language model that can produce coherent and fluent texts on various topics, based on a given input or prompt. It is trained on a large corpus of text data scraped from the web, including books, news articles, social media posts, and more. ChatGPT can also answer questions, write summaries, generate stories, and perform other tasks that involve natural language processing.

ChatGPT uses a deep neural network architecture called Transformer, which allows it to learn complex patterns and relationships among words and sentences. ChatGPT can also adapt to different styles and tones of writing, depending on the context and the domain of the input. For example, ChatGPT can write in a formal or informal way, depending on the situation.

Iowa Schools

How are schools using ChatGPT to ban books?

According to a report by WIRED, some school districts in Iowa have used ChatGPT as a tool to help them comply with SF 496. They have asked ChatGPT questions like “Does [insert book title here] contain a description or depiction of a sex act?” and used its answers as a basis for removing books from their libraries. Some of the books that have been banned by this method include The Handmaid’s Tale by Margaret Atwood, Beloved by Toni Morrison, and The Kite Runner by Khaled Hosseini.

The use of ChatGPT to ban books has raised several concerns among educators, librarians, authors, and students. Some of these concerns are:

The lack of nuance and context in ChatGPT’s answers. ChatGPT may not be able to distinguish between different types and purposes of sexual content in books, such as literary expression, historical accuracy, educational value, or social commentary. ChatGPT may also not be able to account for the age and maturity level of the intended audience of the books.
The potential for bias and error in ChatGPT’s answers. ChatGPT may reflect the biases and prejudices of its training data, which may not be representative of the diversity and complexity of human society and culture. ChatGPT may also make mistakes or give inaccurate or incomplete answers, due to its limitations and uncertainties as an AI system.
The ethical and legal implications of using ChatGPT to censor books. The use of ChatGPT to ban books may violate the intellectual freedom and privacy rights of authors and readers, as well as the academic freedom and professional autonomy of educators and librarians. The use of ChatGPT may also undermine the educational goals and values of schools, such as fostering critical thinking, creativity, curiosity, and empathy among students.

What are the alternatives to using ChatGPT to ban books?

Many experts and advocates have argued that using ChatGPT to ban books is not only ineffective but also harmful for schools and students. They have suggested other ways to deal with the issue of sexual content in books, such as:

Revising or repealing SF 496. Some critics have challenged the constitutionality and validity of SF 496, claiming that it is vague, broad, and arbitrary in its definition and enforcement of sexual content in books. They have called for the revision or repeal of SF 496, or at least for its suspension until further review and clarification.
Using human judgment and expertise. Some proponents have advocated for the use of human judgment and expertise in selecting and evaluating books for school libraries. They have recommended that schools consult with qualified professionals, such as librarians, teachers, parents, students, and experts on literature, education, psychology, and child development. They have also suggested that schools adopt clear and consistent policies and criteria for book selection and review.
Promoting dialogue and education. Some supporters have encouraged schools to promote dialogue and education around sexual content in books, rather than banning them outright. They have advised schools to provide opportunities for students to discuss and learn about sexual issues in a safe and respectful manner, with guidance from trained educators. They have also urged schools to respect the choices and preferences of students and their families regarding what they want to read.

The use of ChatGPT to ban books in Iowa schools has sparked a heated debate over the role of artificial intelligence in education and society. While some see ChatGPT as a convenient and efficient tool to help schools comply with a new law, others see it as a dangerous and irresponsible way to censor books and stifle learning. The controversy highlights the need for more awareness and regulation of the potential benefits and risks of using AI systems like ChatGPT, especially in sensitive and complex domains like education.