Schools have become used to AI tools that work well in short bursts: draft a worksheet, improve an email, summarise a passage, then move on. Anthropic’s latest change points in a different direction. With Claude Opus 4.5, the headline for many schools is Infinite Chats: conversation threads that are effectively unlimited for ongoing work. That sounds simple, but operationally it is a meaningful shift. It changes how staff may organise planning, how students may approach longer projects, and how leaders should think about evidence trails, privacy, and safe use. If your team is already reviewing broader model changes, this briefing sits well alongside our Claude autumn update overview.
On 24 November 2025, Anthropic announced Claude Opus 4.5, with Infinite Chats as one of the most school-relevant practical changes. In plain terms, users can keep a conversation going across much longer periods without treating each task as a fresh start. Instead of repeatedly re-explaining the context for a curriculum unit, a research enquiry, or a drafting process, staff can continue a thread that already holds prior discussion, decisions, and examples.
For schools, the significance is not just convenience. Longer-running threads can support continuity. A head of department might keep one thread for a Year 8 history scheme, revisiting it over several weeks to refine enquiry questions, adapt reading difficulty, and generate extension tasks. A pastoral leader might use a planning thread for assembly themes and communication drafts, provided no personal data is entered. The attraction is obvious: less repetition, more coherence, and faster iteration.
In practice, Infinite Chats means a conversation can act more like a working notebook than a one-off prompt box. The system can draw on the earlier discussion in that thread, so users do not need to restate every assumption, instruction, or example. That can make the output feel more consistent over time. It can also make the process more efficient, especially for complex work that unfolds in stages.
Yet “infinite” should not be read as “use one chat for everything”. Schools should resist that habit early. A very long thread can become muddled, with outdated assumptions mixed into newer instructions. It may also become harder to check where a claim came from, which version was approved, or whether sensitive context slipped in halfway through. This is where operational discipline matters more than the feature itself.
A helpful comparison is with document management. Schools do not put every policy, lesson idea, and safeguarding note into one giant file. They create separate folders and name them clearly. Infinite Chats should be treated in much the same way: useful for continuity within a defined purpose, but not as a catch-all memory space.
Extended student research and project work may be one of the clearest use cases. In a traditional AI workflow, a student researching coastal erosion, renewable energy, or a set text often starts from scratch each time. That can waste time and encourage shallow prompting. With a longer thread, the student can build a sequence: define the question, gather background knowledge, compare viewpoints, draft an outline, revise explanations, and prepare a presentation. The continuity can support better questioning and more purposeful iteration.
Used well, this can improve metacognition. Students can look back at how their thinking changed. A teacher can ask them to identify where the AI response was useful, where it was weak, and what independent checking they completed. That matters because the educational value is not in preserving an endless chat for its own sake. It is in making thinking visible and reviewable.
Still, schools should not mistake continuity for reliability. A long thread can make weak claims feel more persuasive simply because they are repeated. This is one reason evidence routines remain essential. If you are already reviewing how different tools support depth versus speed, our piece on Gemini 3 Flash in classrooms offers a useful contrast in workflow design.
The biggest staff gains are likely to appear in planning, drafting, and iteration. Curriculum planning often happens in layers. A teacher begins with objectives, then decides on prior knowledge, vocabulary, misconceptions, activities, and assessment opportunities. Infinite Chats can support that layered process. Instead of writing a huge prompt once, the teacher can return to the same thread and refine one element at a time.
Consider a science department planning a unit on ecosystems. In week one, the thread helps sequence concepts. In week two, it generates retrieval questions and practical ideas. In week three, it adapts reading tasks for different ages and language levels. The continuity can save time because the model already “knows” the working context of the unit. Similar gains may appear in report comment drafting, policy wording, and communication planning, especially when staff need several rounds of revision. For related audit-trail questions in writing workflows, see our guide to AI-assisted report writing.
There is also a benefit for collaborative professional thinking. A middle leader might use a thread to test several versions of a departmental action plan over time, comparing options before taking a final draft to colleagues. The key is to treat the thread as a drafting partner, not a decision-maker.
Ready to Revolutionise Your Teaching Experience?
Discover the power of Automated Education by joining out community of educators who are reclaiming their time whilst enriching their classrooms. With our intuitive platform, you can automate administrative tasks, personalise student learning, and engage with your class like never before.
Don’t let administrative tasks overshadow your passion for teaching. Sign up today and transform your educational environment with Automated Education.
Very long conversations bring very long shadows. The first risk is unsafe data sharing. As a thread becomes familiar, users may become less cautious and paste in more than they should: pupil names, behaviour notes, SEN details, parent correspondence, or internal safeguarding concerns. That drift is easy to imagine precisely because the chat feels ongoing and trusted.
The second risk is a weak evidence trail. In one long thread, decisions can become buried. Which answer informed the final lesson? Which version was fact-checked? Which prompt generated the wording sent to parents? If schools cannot answer those questions, accountability becomes harder. This is especially important for leaders responsible for policy, data protection, and public communication.
The third risk is over-reliance. A single thread can create the illusion that the model has developed a deep understanding of the school’s needs. In reality, it is still producing responses probabilistically. If staff stop challenging it because “this thread has been right so far”, quality may slip unnoticed. Similar concerns appear in wider discussions about settled practice and assessment integrity, explored in our review of ChatGPT’s current education impact.
SLT, DPOs, DSLs, and IT leads should treat Infinite Chats as a workflow issue, not just a product update. The first question is scope: which use cases are acceptable for long-running threads, and which are not? Curriculum planning may be low risk. Pupil-specific pastoral drafting is not. Schools need that distinction written clearly.
The second question is retention and review. How long should staff keep AI threads? Should important outputs be exported into normal school systems? What must never remain only inside an AI conversation? If a thread informs a policy or a parent communication, there should be a clear route into approved storage and version control.
The third question is training. Staff need examples of safe continuation and safe restart. This is a practical CPD matter, not an abstract policy point. Many schools will want to refresh their AI acceptable use policy and pair that with short training routines, such as those in our INSET day AI workshop guide.
A simple rule helps: continue a chat when the task is the same, the risk is low, and continuity improves quality. Start afresh when the purpose changes, the stakes rise, or the context includes anything sensitive.
So a teacher might continue one thread for developing a geography unit across several weeks. They should start a fresh thread when switching from lesson planning to drafting a letter about a real pupil concern. Likewise, a student might continue a thread for refining an extended project question, but start a new one when beginning an unrelated topic.
Another useful pattern is the checkpoint summary. At key stages, ask the model to summarise the thread’s current decisions, then copy that summary into your own document. This creates a cleaner audit trail and makes it easier to restart elsewhere if the thread becomes cluttered. It also reduces the temptation to keep everything in one place for ever.
When schools test Claude Opus 4.5, keep the evaluation practical. Ask whether Infinite Chats genuinely saves time on extended work. Check whether staff can identify when to restart. Review whether outputs remain easy to verify after several rounds. Test whether the tool encourages better planning habits or simply longer, messier conversations.
It is also worth asking whether your existing rollout model is ready for this kind of feature. Schools that have already set privacy defaults, defined approved uses, and built short training routines will find the transition easier. Our minimum viable AI toolkit guide may help teams that need to tighten those basics first.
Infinite Chats is not just a bigger chat window. It is a shift in how work can accumulate over time. For schools, that opens real opportunities in research, planning, and iterative drafting. But the gains will only be worth having if they come with clear boundaries, strong review habits, and a willingness to start afresh when needed. The most effective schools will not be the ones with the longest threads. They will be the ones with the clearest routines around when a long thread helps and when it quietly becomes a risk.
May your AI workflows stay clear, useful, and well governed.
The Automated Education Team
Microsoft Build 2026 brought a fresh wave of Copilot announcements, but …
Summer term is often the last realistic window for school leaders and AI …
Schools do not need a whole-school AI rollout to learn what works. In many …
Once the papers are marked, many departments want feedback that is sharper …
Primary assessment week can feel intense, especially when SATs and spring …
Exam week often creates pressure to do more, faster, with less time to …
Many school AI rules still rely on vague disclosures such as “I only used …
AI voice tools for modern foreign languages have improved sharply by 2026, …
Spring assessment season puts pressure on teachers, pupils and families to …
Anthropicu Infinite Chats funktsioon muudab seda, kuidas koolid võivad AI-d aja jooksul kasutada, …
Anthropics funktion Infinite Chats förändrar hur skolor kan använda AI över tid, särskilt för längre …
Anthropicin Infinite Chats -ominaisuus muuttaa sitä, miten koulut voivat käyttää AI:ta ajan mittaan, …