Court Backs School's Punishment For Using AI But AI Research Keeps Unveiling New Insights

Every bit of technology comes with its fair share of cynics and critics. Those who warn against the ill effects of overusing technology or having an excessive reliance on it.

Love watching television? It’ll ruin your eyes and rot your brain. Using a refrigerator to store food? You should eat fresh! Using an elevator? You should take the stairs!

Of course, in most cases, the critics have a point. Overdoing anything is harmful to health.

This is one of the major concerns brought up by experts even when it comes to artificial intelligence (AI) and generative artificial intelligence (GenAI).

Other than the obvious threats to national security, jobs, deepfakes, and others that directly influence millions of people, these new technologies have brought in many ethical questions and moral dilemmas, even fatal at times. Especially at a time when businesses are picking up their investments.

Recently, we reported on a student from Hingham High School in Massachusetts getting punished by his school for using AI for his homework.

This led to the parents of the student, Jennifer and Dale Harris, filing a lawsuit in the Plymouth County Superior Court, which was then moved to US District Court for the District of Massachusetts.

They alleged that nowhere in the handbook does any rule restrict the use of AI.

However, the school said that the student admitted to using AI, while the student’s handbook deals with cheating and plagiarism in a section, that bans the “unauthorized use of technology during an assignment”.

Now, an order from the US District Court for the District of Massachusetts ruled in favor of the school, with US Magistrate Judge Paul Levenson noting that the school “have the better of the argument on both the facts and the law.”

The incident, which occurred in December 2023, saw Harris’ son RNH (identified by his initials) and another student (who wasn’t a party to the case) try to pass off GenAI-generated content as their own.

Moreover, they even copied “hallucinated” content, by including citations to nonexistent books.

Despite this, the school allowed the two students to “start from scratch, each working separately, to complete and submit the final project.”

The parents even argued that RNH wasn’t told about the school’s standards for academic honesty as they apply to the use of AI and that the grade should be raised from C-plus to a B.

The school though responded by saying the student “was repeatedly taught the fundamentals of academic integrity, including how to use and cite AI.”


Considering GenAI is a relatively new technology and its use cases are being updated every day, one could wonder if the school’s stance on AI and academic integrity got lost in translation.

Whether or not this is the case with the student, it certainly seems to be the case with the technology, as experts note that AI models work faster when they speak in their own language, instead of translating back and forth in English.

Enter DroidSpeak, a novel framework dedicated to enhancing cross-LLM communication, as per a paper published by Microsoft and the University of Chicago.

The idea is to cut out the bottleneck that builds up between two AI agents when communicating using natural language. When one agent communicates in natural language, it not only sends the current step’s output but also includes the entire conversation history.

The receiving agent must then process the output into a language it understands, to complete its own task.

When this step is repeated several times, it can lead to communication delays and computation overheads. DroidSpeak will help cut the need for translating into natural language and instead directly share the data created in computational steps pre-translation, enabling better scalability and responsiveness, especially in multi-agent systems.

“This strategy circumvents the need for complete context regeneration, enabling faster integration without sacrificing the quality of outputs. By selectively reusing and recalculating only what is necessary, DroidSpeak significantly reduces prefill delays while maintaining high task performance standards,” reads an excerpt from the paper.

Do you think this framework will help businesses boost their productivity? Do you think students should be permitted to freely use AI tools in school projects considering the way the technology is redefining how the world works?

Let us know in the comments below!