Group projects with AI

Dialogia Classroom

See how every student really works with AI.

Classroom turns group projects into transparent work: students use AI under the rules you set, and you see the reasoning, the prompts, and each person's real contribution.

Access Classroom Book a demo

dialogia.org/classroom

Video preview blocked until you accept third-party content.

This is Classroom's full product page. Ready to use it? Sign in or access your account.Open Classroom →

AI work is often invisible

When students use AI privately, teachers see only the final artifact.

Teachers need process evidence

Prompts, drafts, and contributions help explain how learning happened.

Students need guidance, not punishment

Rules and warnings should support better work, not frame AI as surveillance.

Integration, not just a tool

AI joins the group as one more teammate — and everything stays visible to the teacher. That's what we mean by integrating AI into the classroom.

Review the AI chat

See and review each student's full conversation with the AI, not just the final deliverable.

Assess the process, not only the result

Evaluate how the work was done — the thinking, the iterations — alongside the outcome.

Real contribution per student

In group work, see who contributed what, and how much of the content was AI vs the student's own.

AI as one more teammate

Students collaborate with each other and the AI joins as another teammate, assisting the group.

The teacher joins in

Step into the conversations yourself to guide, nudge, or unblock the group when needed.

How Classroom works

The teacher decides how much AI the student can use.

Each activity launches with a work mode. The AI adapts its behavior and assessment takes the rules into account.

🤝

Assisted

“Co-write with me”

Students can lean on the AI freely — it can generate full paragraphs or complete drafts. They supervise, critique, and decide.

Default

⚖️

Balanced

“Let's think together”

The default. The AI is a thinking partner: it suggests, gives feedback, helps structure, and rewrites specific fragments on request.

✏️

Strict

“Only if I ask”

Ideas and writing are 100% the student's. The AI only helps with spelling, grammar, synonyms, or paraphrasing what they've already written.

🏛️

Socratic

“Discover it yourself”

The AI never gives the answer directly. Instead, it asks questions that help the student think, reflect, and reach their own conclusions.

Per activity, the teacher can also let students attach their own documents (PDF, DOCX…) and enable web search with cited sources.

Built-in guardrail

When a student tries to shortcut the thinking, everyone sees it.

If a message goes against the activity's mode — like asking the AI to “just give me the answer” in Socratic mode — Dialogia flags it before it's sent, explains why, and logs the attempt for the teacher. The AI still won't hand over the answer. Misuse becomes visible, not invisible.

A real-time nudge back on task — the student can rephrase or send anyway.
Every attempt is logged and attributed to the student.
The model holds the line: in Socratic mode it keeps asking, never answers.

Ethics of AI · class debateSocratic mode

Should our school use facial recognition to take attendance? Take a position and defend it.

This message may go against the activity's mode

You can send it anyway: it will be logged for your teacher, and the AI will keep answering with questions, not with conclusions.

You are asking the AI to decide for you and to set the limits of the debate. In Socratic mode it is you who builds the arguments and reaches the conclusion.

Your message

Just tell me whether it is right or wrong, give me the three best arguments and where to draw the line, so I can write my essay.

Cancel and rephraseSend anyway

Socratic mode — the warning a student sees before sending an off-mode message.

Real participation %

Active time, prompts made, contribution in shared vs private chat.

Prompt traceability

Every interaction with the AI is logged and attributed to the student who made it.

Evidence-based assessment

Automatic rubrics that cross the final work with the individual process.

Group projects with AI

How a student works with the AI

Balanced mode, live: the AI guides instead of solving — and every step stays visible to the teacher.

dialogia.org/classroom/trabajo

Balanced

Your Work

Write a financial viability analysis for a startup of your choice.

Healthy-food delivery app (Madrid)

1) Target market: urban 25–45, want to eat well without cooking.
2) Revenue: 20% commission + €9.99/mo premium.
4) Break-even: €18 ticket · €3.6 margin/order · €4,000 fixed → ~1,111 orders/mo (≈37/day).

AI assistant

How should I structure a financial viability analysis for my app?

Let's build it together. Start with market & size, revenue model, costs and break-even. Who exactly is your app for?

The AI guides — it doesn't do it for you.

Can you search and cite the market size for me?

Guardrail

This may go against the activity's mode: you're asking the AI to do the core work you should do yourself. It will still answer, but this is logged.

Guardrail: it warns the student — and records it.

Good start. Your €18 ticket is reasonable (Statista 2023), but you're leaving out payment fees and seasonality.

Feedback on the student's own work — the process is assessable.

Group projects with AI

What the teacher sees

The same activity, from the other side: the whole process is visible, and the AI grades by evidence — taking the work mode into account.

dialogia.org/classroom/actividad

AlexCompleted

4 min active

Process visibility

Full chat with the AI (4 prompts, timestamped)
Active time per student
Real contribution: AI vs the student's own

AI evaluation by evidence

8/10

The grade respects Balanced mode: active, critical use of the AI is rewarded.

Quality of AI use5/5
Coherence & structure4/5
Argument & depth3/5
Language5/5

“Doesn't go deeper into the risks the AI flagged in the chat.” — the evaluation cites the conversation as evidence.

For professors and institutions

Activity modes and rules
Teacher-guided AI expectations
Group work and individual contribution signals
Evidence for feedback and evaluation

For students

Clear guidance before using AI
A transparent process that explains expectations
Fairer assessment of work and contribution
Support without punitive surveillance framing

How it works

Teacher defines the task and AI mode

The activity starts with clear rules: how AI may help, what students must own, and what evidence matters.

Students work with a visible process

Prompts, drafts, warnings, and contributions stay available as learning evidence, not hidden shortcuts.

Teacher reviews evidence and learning

Assessment can consider the final work and the process that produced it.

Built on the Dialogia pillars

Precision RAG

The AI answers only with the materials the teacher validates. It cites the page and slide. If it's not in the syllabus, it doesn't make it up.

Visible pedagogy

We turn private interactions into pedagogical analytics: which topics confuse, which slides fail, who contributes what.

Data sovereignty

European cloud, GDPR by contract, guaranteed no-training. The institution owns the intelligence it generates.

Transparency that supports learning.

Dialogia is designed around governed AI use: teacher-defined materials and rules, transparent participation modes, and privacy-aware records that support learning rather than surveillance. In Classroom, process visibility is framed as pedagogical evidence.

Teacher-defined rulesProcess visibilityPrivacy-aware recordsNot surveillance

Bring governed AI into your classroom.

Create your first activity in minutes, or book a demo and we'll walk you through it.

Access Classroom Book a demo