One Login, Five Names on the Rubric: How Group Projects Got Outsourced to a Chatbot
Students split the cost of a single ChatGPT account and run whole projects through it. Banning the tool won't teach anyone to write a paragraph on their own.
A ChatGPT Plus login runs about ₱1,000 a month in the Philippines, roughly US$18 with local VAT. Split five ways, that's about ₱200 a head. OpenAI's cheaper Go tier lands at ₱300, and either way the math works out to less than a week of coffee runs. Plenty of students run their group work off a login that lives in a Google Doc titled "DO NOT SHARE."
The account writes the intro. It builds the outline, drafts the body, and cleans up the citations someone half-remembered. One person pastes it into the template. Everyone else confirms they've read it in the group chat, and the deliverable ships before midnight.
The tool is not the cheat. The blank page is.
Here's the part nobody wants to say out loud: a lot of students reaching for autocomplete never got much practice writing without it. Educators have spent years warning that pandemic-era distance learning left gaps in exactly the skills a chatbot now papers over, arguing a point, building a case across paragraphs, drafting from nothing. The muscle for that was never fully trained, so of course the reflex is to outsource it.
Lecturers see the output, not the gap. So the response is a ban. No AI, honor pledge signed, laptops closed, blue books out. The problem is that a ban tests a skill the student may never have been given time to build. You can't fail your way into fluency you were never taught.
Banning it just moves the login off campus
Enforcement here is theater. AI-detection software is documented as unreliable: multiple education studies have found the tools produce both false positives and false negatives, and non-native English writers get flagged at higher rates. OpenAI pulled its own detection classifier back in 2023, citing low accuracy. Faculty can't reliably tell machine writing from human writing, and students know it.
A ban doesn't remove the login. It teaches everyone to hide it better, to add typos, to run the draft through a paraphraser so the flag stays green.
Meanwhile the students who genuinely want to learn get punished twice. They lose the tool that could scaffold a first draft, and they still don't get taught the thing the tool replaced. The ones gaming it lose nothing. They were never going to write it themselves.
What the classroom owes them instead
The honest move is to teach in front of the tool, not against it. Make students defend a paragraph out loud. Have them mark what the chatbot got wrong, because it gets plenty wrong, and grade the correction. Draft in class, on paper, where the account can't reach. Show them what a bad AI paragraph looks like so they can smell one at a job later.
None of this is a ban and none of it is surrender. It's the work of actually teaching people to think on a page, which is harder than writing a syllabus line that says "AI is prohibited."
The shared login isn't going anywhere. It costs less than the group's coffee tab and it already knows the assignment. The question on the table is whether a student walks out of four years able to write a paragraph the account didn't touch, or whether the diploma just certifies that they knew where the Google Doc lived.
