Incident in Alaska Highlights Challenges of AI Use in Educational Policy Making
A recent incident in Alaska has brought to light the challenges of using generative artificial intelligence (AI) in educational administration. The state's Department of Education released a policy document that included multiple inaccurate citations, attributed to the use of AI in its development. This incident has raised questions about the reliability of AI-generated content and the necessity for rigorous human oversight in educational policy creation.
The incident involved the drafting of a proposed policy on mobile phone use in Alaskan schools. The document, crafted under the guidance of Education Commissioner Deena Bishop, was circulated ahead of a state board meeting. It referenced several studies purportedly supporting the department's stance, including a study supposedly published in the journal Computers in Human Behavior, and others from the American Psychological Association and the Journal of Educational Psychology. However, it was discovered that four out of the six citations were fictitious.
Upon investigation by the Alaska Beacon, it was revealed that the studies had never been published, nor could their titles be located through extensive online searches. Hyperlinks accompanying the citations led users to unrelated content, such as analyses on "Sexualized Behaviors on Facebook," bearing no relevance to mobile phone usage in school environments.
The discrepancies prompted queries from the media, leading the department to revise the online document initially. A department spokesperson described these citations as "placeholders," supposedly included during the drafting phase before replacing them with accurate information. Commissioner Bishop admitted to using a generative AI platform for drafting the resolution, which contributed to the citation errors. Although Bishop claimed to have identified and corrected these issues before the board meeting, remnants of the AI-induced errors persisted in the updated document.
Generative AI platforms are known for "hallucinations"—the production of seemingly credible but incorrect information. Such hallucinations occur due to the design of AI systems that generate responses based on patterns learned from vast datasets, optimised for coherence rather than factual precision. This architectural flaw came to public notice last year when two New York lawyers faced sanctions for submitting a legal brief featuring fabricated legal citations by an AI model.
In light of these events, it has become imperative for educational institutions to develop strategies to mitigate the risks associated with AI usage. Schools and educational bodies must ensure the integrity and accuracy of data underpinning policy and administrative documentation. Experts suggest a multi-faceted approach, involving the establishment of clear AI usage guidelines, mandatory human fact-checking, and comprehensive AI literacy training for staff. Additionally, emphasising the use of peer-reviewed, human-curated sources can further safeguard against misinformation.
This incident serves as a pertinent example of the complexities and potential pitfalls in integrating AI into administrative tasks within education. While AI has the capacity to streamline certain processes, balancing its use with stringent oversight is crucial to maintain the credibility and reliability of educational policies.
Source: Noah Wire Services