{"course":{"slug":"ai-for-students","title":"AI for Students","url":"https://addaly.com/learn/ai-for-students","level":"beginner","track":"work"},"lesson":{"slug":"group-work-and-shared-rules","title":"Group work, and shared rules","url":"https://addaly.com/learn/ai-for-students/group-work-and-shared-rules","position":55,"of":73,"minutes":7,"answer":"Agree the group's AI rule in writing before the first session, work in a shared document with version history, and require each person to confirm they opened every source in their own section.","keyIdea":"Agree the group's AI rule in writing before the first session, work in a shared document with version history, and require each person to confirm they opened every source in their own section."},"body":{"format":"markdown","text":"## The submission is joint and the consequences are not\n\nWhen four people submit one document, all four are answerable for all of it. If one section contains a fabricated citation or generated prose that the group's declaration did not cover, the investigation involves everybody, and *I did not know* is a weak position that also damages a friendship.\n\nThis is the most common way careful students end up in integrity meetings, and it is almost entirely preventable by a fifteen-minute conversation nobody wants to have at the start.\n\n## The conversation, in one message\n\nSend it before the first working session:\n\n> Before we start: how are we using AI on this? My suggestion: fine for generating questions, critique of our own drafts, translation, and checking understanding. Not for prose that goes in the document. Everyone opens every source they cite. Everyone writes their own section. If anyone wants to work differently, say now rather than at the end.\n\nAwkward for about ninety seconds. Then it is settled, in writing, with a timestamp, and everybody knows the standard.\n\nWhat matters is not that everybody adopts the strictest possible rule. It is that the rule is explicit and shared. A group where two people think generated drafting is fine and two think it is cheating will have a problem, and it will arrive at the worst moment.\n\n## Check the actual policy first\n\nCourse policies on group work sometimes differ from individual-work policies, and some require a group declaration listing what each member did and what tools were used. Read yours before agreeing anything, so the group rule is at least as strict as the institutional one. If your policy is silent on group work, ask the module leader in writing: the reply is useful evidence later.\n\n## Working in a shared document\n\nPractical arrangements that prevent most problems:\n\n- **Google Docs or Word online**, with version history on. Every contribution is attributed and timestamped automatically. This is the single most useful thing you can do, and it costs nothing.\n- **Never paste a finished block into a shared document.** Write in it, or paste in pieces with edits, so the history shows work rather than an arrival. A large block appearing at once is what triggers questions, whoever wrote it.\n- **One person owns each section**, named at the top of the draft.\n- **One person does a consistency pass at the end**, and says so.\n\n## The reference list is where it goes wrong\n\nCombining four bibliographies is the moment fabricated citations enter a document, because the person merging did not read anybody else's sources and assumes they were checked.\n\nThe rule that fixes it: **whoever cites it, opens it, and says so.** Before submission, each person confirms in the group chat that every reference in their section was opened and that they can say which page they used. That message is your record, and it takes two minutes.\n\nIf someone will not confirm, the reference comes out. This is not distrust; it is exactly the standard everybody would want applied to their own name on the cover.\n\n## Dividing the work, revisited\n\nFrom the study-groups lesson, but with the assessment consequence attached. Splitting topics so each person only learns a quarter is efficient for producing a document and poor for learning, and in any group assessment with an individual component (a viva, a reflective piece, a question in an exam) it is directly costly.\n\nThe arrangement that survives: split the *writing*, share the *understanding*. Everybody reads the whole draft, and everybody should be able to defend any section. Half an hour before submission, take turns explaining each other's sections. It is the fastest possible check for anything nobody actually understands, which is also the fastest possible check for anything nobody actually wrote.\n\n## If it goes wrong\n\nIf you find, late, that a section contains work you are not comfortable submitting: raise it with the group first, immediately, in writing. If it cannot be resolved, tell the module leader before submission rather than after. Disclosing early is treated completely differently from being found later, at every institution, and it is also the only version you will feel fine about afterwards."},"check":{"question":"Why is \"whoever cites it, opens it, and confirms so before submission\" the specific rule that protects a group?","options":[{"text":"Because it distributes the checking workload evenly, which makes verification feasible for a long reference list.","correct":false,"tempting":"Frames it as efficiency; the rule would still be necessary for a list one person could check alone."},{"text":"Because institutional policies on group work generally require an individual declaration from each member about the sources they personally consulted.","correct":false,"tempting":"Invents a common requirement; declarations vary widely and many groups submit none, while the underlying hazard is present either way."},{"text":"Because it creates a written record that identifies who is responsible if a fabricated reference is found.","correct":false,"tempting":"Treats the rule as apportioning blame after the fact; joint submissions implicate everyone regardless of who wrote what."},{"text":"Because merged bibliographies are where unchecked references enter, since the person combining them has not read anyone else's sources.","correct":true}]},"prev":{"slug":"presentations-and-visuals","title":"Presentations, posters and visuals","url":"https://addaly.com/learn/ai-for-students/presentations-and-visuals"},"next":{"slug":"the-rules-and-the-detectors","title":"The Rules, And Why Detectors Do Not Work","url":"https://addaly.com/learn/ai-for-students/the-rules-and-the-detectors"},"authorship":"Founder and series editor: Dr Mustafa Mun. Written by AI models to a written standard. Exam keys checked blind by a second model: 3,665 of 3,666 agreed. Academic review invited.","review":{"status":"Draft, open for academic review"},"cite":"Addaly (2026) 'Group work, and shared rules', in AI for Students. Edited by M. Mun. Addaly. Available at: https://addaly.com/learn/ai-for-students/group-work-and-shared-rules (Accessed: 13 October 2026).","licence":{"name":"CC BY-NC-SA 4.0","url":"https://creativecommons.org/licenses/by-nc-sa/4.0/"},"reader":{"schema":1,"title":"Group work, and shared rules","courseTitle":"AI for Students","module":{"index":6,"slug":"writing-and-voice","title":"Writing, and keeping your voice"},"answer":null,"keyIdea":"Agree the group’s AI rule in writing before the first session, work in a shared document with version history, and require each person to confirm they opened every source in their own section.","hero":{"kind":"plate","image":null,"alt":null,"caption":null},"blocks":[{"t":"h2","id":"the-submission-is-joint-and-the-consequences-are-not","runs":[{"text":"The submission is joint and the consequences are not"}]},{"t":"p","runs":[{"text":"When four people submit one document, all four are answerable for all of it. If one section contains a fabricated citation or generated prose that the group’s declaration did not cover, the investigation involves everybody, and "},{"text":"I did not know","i":true},{"text":" is a weak position that also damages a friendship."}]},{"t":"p","runs":[{"text":"This is the most common way careful students end up in integrity meetings, and it is almost entirely preventable by a fifteen-minute conversation nobody wants to have at the start."}]},{"t":"h2","id":"the-conversation-in-one-message","runs":[{"text":"The conversation, in one message"}]},{"t":"p","runs":[{"text":"Send it before the first working session:"}]},{"t":"quote","runs":[{"text":"Before we start: how are we using AI on this? My suggestion: fine for generating questions, critique of our own drafts, translation, and checking understanding. Not for prose that goes in the document. Everyone opens every source they cite. Everyone writes their own section. If anyone wants to work differently, say now rather than at the end."}]},{"t":"p","runs":[{"text":"Awkward for about ninety seconds. Then it is settled, in writing, with a timestamp, and everybody knows the standard."}]},{"t":"p","runs":[{"text":"What matters is not that everybody adopts the strictest possible rule. It is that the rule is explicit and shared. A group where two people think generated drafting is fine and two think it is cheating will have a problem, and it will arrive at the worst moment."}]},{"t":"h2","id":"check-the-actual-policy-first","runs":[{"text":"Check the actual policy first"}]},{"t":"p","runs":[{"text":"Course policies on group work sometimes differ from individual-work policies, and some require a group declaration listing what each member did and what tools were used. Read yours before agreeing anything, so the group rule is at least as strict as the institutional one. If your policy is silent on group work, ask the module leader in writing: the reply is useful evidence later."}]},{"t":"h2","id":"working-in-a-shared-document","runs":[{"text":"Working in a shared document"}]},{"t":"p","runs":[{"text":"Practical arrangements that prevent most problems:"}]},{"t":"ul","items":[[{"text":"Google Docs or Word online","b":true},{"text":", with version history on. Every contribution is attributed and timestamped automatically. This is the single most useful thing you can do, and it costs nothing."}],[{"text":"Never paste a finished block into a shared document.","b":true},{"text":" Write in it, or paste in pieces with edits, so the history shows work rather than an arrival. A large block appearing at once is what triggers questions, whoever wrote it."}],[{"text":"One person owns each section","b":true},{"text":", named at the top of the draft."}],[{"text":"One person does a consistency pass at the end","b":true},{"text":", and says so."}]]},{"t":"h2","id":"the-reference-list-is-where-it-goes-wrong","runs":[{"text":"The reference list is where it goes wrong"}]},{"t":"p","runs":[{"text":"Combining four bibliographies is the moment fabricated citations enter a document, because the person merging did not read anybody else’s sources and assumes they were checked."}]},{"t":"p","runs":[{"text":"The rule that fixes it: "},{"text":"whoever cites it, opens it, and says so.","b":true},{"text":" Before submission, each person confirms in the group chat that every reference in their section was opened and that they can say which page they used. That message is your record, and it takes two minutes."}]},{"t":"p","runs":[{"text":"If someone will not confirm, the reference comes out. This is not distrust; it is exactly the standard everybody would want applied to their own name on the cover."}]},{"t":"h2","id":"dividing-the-work-revisited","runs":[{"text":"Dividing the work, revisited"}]},{"t":"p","runs":[{"text":"From the study-groups lesson, but with the assessment consequence attached. Splitting topics so each person only learns a quarter is efficient for producing a document and poor for learning, and in any group assessment with an individual component (a viva, a reflective piece, a question in an exam) it is directly costly."}]},{"t":"p","runs":[{"text":"The arrangement that survives: split the "},{"text":"writing","i":true},{"text":", share the "},{"text":"understanding","i":true},{"text":". Everybody reads the whole draft, and everybody should be able to defend any section. Half an hour before submission, take turns explaining each other’s sections. It is the fastest possible check for anything nobody actually understands, which is also the fastest possible check for anything nobody actually wrote."}]},{"t":"h2","id":"if-it-goes-wrong","runs":[{"text":"If it goes wrong"}]},{"t":"p","runs":[{"text":"If you find, late, that a section contains work you are not comfortable submitting: raise it with the group first, immediately, in writing. If it cannot be resolved, tell the module leader before submission rather than after. Disclosing early is treated completely differently from being found later, at every institution, and it is also the only version you will feel fine about afterwards."}]}],"check":{"question":[{"text":"Why is “whoever cites it, opens it, and confirms so before submission” the specific rule that protects a group?"}],"options":[{"text":[{"text":"Because it distributes the checking workload evenly, which makes verification feasible for a long reference list."}],"correct":false,"tempting":[{"text":"Frames it as efficiency; the rule would still be necessary for a list one person could check alone."}]},{"text":[{"text":"Because institutional policies on group work generally require an individual declaration from each member about the sources they personally consulted."}],"correct":false,"tempting":[{"text":"Invents a common requirement; declarations vary widely and many groups submit none, while the underlying hazard is present either way."}]},{"text":[{"text":"Because it creates a written record that identifies who is responsible if a fabricated reference is found."}],"correct":false,"tempting":[{"text":"Treats the rule as apportioning blame after the fact; joint submissions implicate everyone regardless of who wrote what."}]},{"text":[{"text":"Because merged bibliographies are where unchecked references enter, since the person combining them has not read anyone else’s sources."}],"correct":true,"tempting":null}],"order":[0,2,1,3],"why":null},"prev":{"slug":"presentations-and-visuals","title":"Presentations, posters and visuals","minutes":8},"next":{"slug":"the-rules-and-the-detectors","title":"The Rules, And Why Detectors Do Not Work","minutes":9}}}