IB Internal Assessment AI Tutor Playbook 2026

Audience: IB Diploma Programme students (Year 12 / Year 13 / DP1 / DP2) entering the 2027 May exam session, their parents, and IA subject teachers who want a rubric-aware workflow. Covers the seven most common IA subjects (Biology, Chemistry, Physics, History, English, Maths AA/AI, Psychology, Economics, Geography) plus a general method that maps to any IA in any subject group.

Hook: Every IB Diploma student writes at least one Internal Assessment per subject — and for SL students that is one IA per subject, for HL students that is one or two IAs per subject, meaning the typical IB Diploma cohort writes 6 to 10 IAs across the two-year programme. The IA is 20 to 25 percent of the final subject grade for every IB subject except the arts and language acquisition. It is graded by the classroom teacher and externally moderated by IB examiners who spot-check 5 samples per subject per school. The total rubric is 22 marks (criteria vary by subject but the A-to-E band descriptors stay the same), and the grade C killer is almost always the same pair of criteria — Criterion C (Critical Thinking / Analysis & Evaluation) and Criterion D (Reflection). An AI tutor that knows the per-subject rubric can stress-test an IA draft paragraph against the 3/3 band descriptors, roleplay the IB moderator question ("show me how this evaluation addresses the source"), and walk a student from blank page to grade 7 IA in 6 to 10 weeks. This is the workflow, subject by subject, with the AI tutor prompts that mirror IB examiner moderation.

Tone: Examiner-aware, rubrical, per-subject specificity. For students who have a topic and need to convert it into a grade 7.

Word count target: 3,500–4,000

---

Why the IA is the most coachable 20% of your IB grade

The IA is the IB component that is most under direct student control. The external exam is fixed by the question paper — the student gets one shot in May 2027 and either knows the content or does not. The IA is a piece of work the student drafts, gets feedback on, redrafts, and submits on a date that is internal to the school (typically March of Year 13 for HL subjects, January of Year 13 for SL subjects in the May 2027 session). That means the IA is a 6-to-10-week open-book project in which the student can have a teacher mark three drafts, an AI tutor mark a fourth draft against the rubric, and a supervisor sign off before submission. No other IB component gives that latitude.

The leverage is also huge. An IA that lifts from grade 6 to grade 7 lifts the subject grade by 1 IB point (out of 7). Across 6 IAs, that is potentially 6 IB points — which is the difference between a 34-point diploma (top 30% of IB cohort) and a 40-point diploma (top 5%, automatic offer from Oxbridge / Ivy / NUS / UCL). The same student, the same exam preparation, but a 6-point IA lift, walks into a different university tier.

The 2025 cohort moderation data tells the same story it has told since 2018: the modal IA grade worldwide is 5 (out of 7), and the failure rate (grade 1–2) sits around 9 percent. The grade 5 modal is not because the median student cannot produce a grade 7 IA. It is because the median IA does two things wrong:

1. Criterion C (Critical Thinking / Analysis & Evaluation, typically 0–3 or 0–4 marks) is under-developed. The student describes what they did, shows the data, but never argues with the data. They treat the IA as a lab report or a research summary rather than an argument.
2. Criterion D (Reflection, typically 0–3 marks) is generic. The student writes "I learned a lot from this IA and I now understand the topic better" instead of "My initial hypothesis was that X, but the deviation of 14% from the expected value at trial 4 forces me to reconsider the assumption that Y is linear in this concentration range, and the second iteration of the methodology corrected for this by Z."

An AI tutor that has been prompt-engineered to mark against the per-subject IB rubric — including the 2024-onwards moderation emphasis on subject-specific critical thinking — closes both gaps.

---

The 22-point IB IA rubric: shared architecture across subjects

Although every IA subject has its own criterion names and mark allocations, the architecture is consistent across the IB Diploma. There are five bands of performance: 0 (no evidence), 1–2 (limited), 3–4 (adequate), 5–6 (substantial), 7 (excellent). And the criteria cluster around the same four skills:

Skill	What it measures	Typical criterion name
Criterion A (focus / research question / engagement with literary text)	Does the student have a precise, defensible, appropriately-scoped question or focus?	"Focus", "Engagement", "Research Question", "Engagement with Literary Text"
Criterion B (knowledge, understanding, methodology, application of literary features)	Does the student demonstrate subject-specific knowledge and use the right methodology or analytical tools?	"Knowledge & Understanding", "Methodology", "Application of Analytical Skills"
Criterion C (critical thinking, analysis, evaluation, personal engagement for the EE)	Does the student argue with the evidence rather than merely describe it?	"Critical Thinking", "Analysis & Evaluation", "Critical Investigation"
Criterion D (reflection, conclusion, presentation for some subjects)	Does the student reflect on what they learned and how their thinking evolved?	"Reflection", "Conclusion", "Reflection & Evaluation"
Criterion E (some subjects add a fifth for formal conventions / structure / bibliography)	Is the IA correctly formatted and structured?	"Formal Presentation", "Structure", "Bibliography", "Format"

The total marks across these criteria sum to 22 in most IB subjects (Biology: 24, Chemistry: 24, Physics: 24, History: 25, English A: 25, Maths AA/AI: 20, Psychology: 22, Economics: 25, Geography: 25). After the raw marks are summed, IB converts to a 1–7 scale using subject-specific grade boundaries that are not published until results day in July.

The single most important fact about this architecture: Criterion C and Criterion D together are 6–8 marks out of 22–25. That is, on a 22-mark rubric, C + D is up to 36 percent of the grade. A student who is solid on A and B (research question, methodology) but weak on C and D (analysis, reflection) caps their IA at grade 5. To break through to grade 7, the student must score 5–6 of the available 6–8 marks on C and D combined.

This is why the AI tutor workflow below prioritises C and D above everything else.

---

The 6-month IA workflow that an AI tutor can run per subject

The IA calendar for a May 2027 exam session is roughly:

• September 2026 (DP1 / Year 12 start): IA topic selection and research question draft
• October–December 2026: methodology / data collection / reading list assembly
• January–February 2027 (DP2 / Year 13 start): first full draft
• March 2027: teacher-marked second draft + student redraft
• April 2027: final submission + teacher mark + upload to IB

That gives a typical IA a 6–8 month production timeline, with 3 to 4 weeks of the most intensive work concentrated in the redraft-and-finalise window. The AI tutor workflow plugs into each stage as follows:

Stage 1 — Topic and Research Question (4–6 weeks, Sep–Oct). AI tutor role: stress-test the research question against the scope criterion. The student proposes a question, the AI tutor asks: "If a student did exactly this in 4,000 words, would the IB moderator say the scope is too broad, too narrow, or appropriately focused? What is the dependent variable? What is the independent variable? What is the controlled variable? Is the question answerable in the time available?" For Biology, Chemistry, and Physics IAs the AI tutor should ask for the variables explicitly; for History, English, and Economics IAs the AI tutor should ask for the source list and the thesis sentence.

Stage 2 — Methodology and Sources (6–10 weeks, Oct–Dec). AI tutor role: flag methodology risks the student has not seen. For Biology IAs, this is where the biggest grade C killer lives — a methodology that does not control for the most obvious confounding variable. For History, this is where the secondary-vs-primary source balance goes wrong. For English A IAs, this is where the body of the essay becomes description of literary features rather than argument about their effect. The AI tutor prompts for each are below.

Stage 3 — First Draft (4–6 weeks, Jan–Feb). AI tutor role: mark the draft against the per-subject rubric. The student pastes the draft into the AI tutor with the prompt: "Mark this IA draft against the IB [subject] IA rubric, criterion by criterion, giving a mark out of [criterion marks] and a 2-sentence justification. End with a list of the 3 changes that would lift the mark by the most." This is the rubric-aware marking pass that the classroom teacher will do in March, except the AI tutor does it in February so the student has a month to redraft.

Stage 4 — Redraft (3–4 weeks, March). AI tutor role: re-mark the redraft against the same rubric, compare to Stage 3 marks, and identify any criterion that has not lifted. The teacher does the final human mark in late March, but the AI tutor has pre-empted the C and D gap before the teacher sees the work.

Stage 5 — Submission (April). AI tutor role: format and citation check. Bibliography in MLA / APA / Chicago as required by the subject guide, figures numbered and captioned, appendix included if relevant, word count within the subject limit.

The remainder of this playbook gives the subject-specific AI tutor prompts for the seven most common IA subjects.

---

IB Biology IA: the lab report with critical thinking in Criterion C

The Biology IA is a 6–12 page lab report on an investigation the student designs, executes, and writes up. It is graded on five criteria totalling 24 marks: Personal Engagement (0–2), Exploration (0–6), Analysis (0–6), Evaluation (0–6), Communication (0–4). The grade 7 modal IA scores 22/24. The grade 5 modal IA scores 16–18/24, and the loss is almost always concentrated in Evaluation (0–6) — the criterion that asks the student to evaluate the methodology, the sources of error, and the extent to which the conclusion is supported.

The single biggest grade C killer in Biology IA: the conclusion repeats the hypothesis and the data summary without evaluating the methodology. A typical grade 5 conclusion reads "The data supports the hypothesis. The trend was as expected. This shows the relationship between X and Y." A grade 7 conclusion reads "The data partially supports the hypothesis — the trend was in the predicted direction at low concentrations (0–2 M) but reversed at high concentrations (>3 M), which suggests that the rate-limiting step in the reaction is concentration-dependent in a non-linear way and that the assumption of a simple first-order kinetic model is only valid in the lower concentration range. The most significant source of error is the temperature drift of ±1.5°C across the 8 trials, which is comparable in magnitude to the effect being measured and would need to be controlled with a water bath in any future iteration."

AI tutor prompt for the Biology IA draft:

"Mark this Biology IA draft against the IB Biology IA rubric, criterion by criterion. For each criterion, give a mark out of the available marks and a 2-sentence justification quoting the specific paragraph from the IA that supports your mark. End with: (a) the 3 highest-leverage changes the student should make in the redraft, and (b) whether the Evaluation section would survive a strict IB moderator who looks for explicit comparison between the data and the hypothesis, identification of specific sources of error with quantitative reasoning, and suggestions for improvement that are methodologically realistic (not 'I would do more trials' but 'I would control for temperature using a thermostatted water bath at 25.0 ± 0.1°C')."

Internal links to IB Biology IA workflow:

• → /blog/ai-tutor-ib-extended-essay-playbook-2026 — post-62, the EE workflow that the Biology IA methodology mirrors
• → /blog/ai-tutor-ks3-year-9-options-ib-myp-2026 — post-35, KS3 MYP to IB Biology transition

---

IB Chemistry IA: the same lab-report structure, the same C and D killers

The Chemistry IA rubric is structurally identical to Biology (Personal Engagement, Exploration, Analysis, Evaluation, Communication, totalling 24 marks). The differences are subject-specific:

• The IA must be on a chemistry topic — not biology, not physics, not interdisciplinary.
• The methodology must include at least one quantitative measurement with units and uncertainty propagation.
• The Evaluation must address chemical-specific sources of error (calibration of glassware, purity of reagents, side reactions, equilibrium assumptions).

The single biggest grade C killer in Chemistry IA: the Evaluation section lists sources of error without quantifying them. A grade 5 Evaluation reads "There may have been impurities in the reagents and the glassware may not have been perfectly clean." A grade 7 Evaluation reads "The 2.3% deviation between the measured and theoretical yield for trial 3 is most likely attributable to the 0.05 g of solid residue recovered from the filter paper, which corresponds to 4.2% of the limiting reagent — this is within the cumulative uncertainty of the balance (±0.01 g) and the filter paper mass transfer (~1%), so the deviation is consistent with experimental error rather than a flaw in the procedure."

AI tutor prompt for the Chemistry IA draft: same as the Biology prompt, but with the subject name swapped and the rubric references updated. The AI tutor should also be prompted to "specifically flag if the Evaluation section names any source of error without a quantitative estimate of its magnitude" — that is the single most common grade 5 trap.

---

IB Physics IA: design before data

The Physics IA is graded on the same five criteria (24 marks total), but the workflow is inverted: the student must complete the Design stage of the IA before the laboratory work begins, and the Design criterion (part of Exploration, 0–6) carries the most marks.

The single biggest grade C killer in Physics IA: the student designs the experiment after doing it. The IB moderator can tell, because the Design section reads like a description of what happened rather than a plan for what was about to happen, and there is no evidence of the student iterating the design after a pilot trial. The grade 7 IA includes a 1-page Design section that explicitly identifies the independent variable, the dependent variable, the controlled variables, the measurement instrument for each variable with uncertainty, and the methodology for at least 5 trials across a range of values.

AI tutor prompt for the Physics IA Design section:

"Read this Physics IA Design section and check whether it satisfies the IB Physics rubric requirement for the Design component of Exploration: (1) is the research question stated as a clear, testable hypothesis with the dependent and independent variables identified? (2) are the controlled variables listed with the specific way they will be controlled? (3) is the measurement uncertainty for each variable specified? (4) is the range of the independent variable specified, with at least 5 trial points? (5) is there evidence the design was iterated based on a pilot trial? Score the Design section out of the available marks (typically 4–5) and identify the single highest-leverage gap."

---

IB History IA: the 2,200-word investigation that is half source evaluation

The History IA is a 2,200-word historical investigation on a topic of the student's choice. It is graded on five criteria totalling 25 marks: Identification and Evaluation of Sources (0–5), Investigation (0–15), Reflection (0–5). Note the weighting: the Investigation criterion alone is 15/25 = 60 percent of the grade, and the Reflection criterion is 5/25 — meaning the grade is dominated by whether the student argues with sources, not just describes them.

The single biggest grade C killer in History IA: the body of the investigation summarises what historians have said rather than arguing with them. The grade 5 IA reads "Historian A argues X. Historian B argues Y. Historian C argues Z." The grade 7 IA reads "Historian A's argument is convincing on the political dimension but fails to account for the economic data published by the Bank of England in 1928, which suggests that the 1926 General Strike had a measurably larger impact on industrial output than A acknowledges. The implication is that A's thesis applies to political effects but not economic effects, and the investigation below tests this claim using the previously-uncited Bank of England archive material."

AI tutor prompt for the History IA:

"Mark this History IA against the IB History IA rubric. For each of the three criteria (Identification and Evaluation of Sources, Investigation, Reflection) give a mark out of the available marks and a 2-sentence justification. Specifically flag: (a) whether the Investigation section argues with at least 3 historians rather than merely summarising them, (b) whether the Evaluation of Sources section applies the OPVL framework (Origin, Purpose, Value, Limitation) to each of the student's 2 primary sources, and (c) whether the Reflection section identifies a specific way the investigation changed the student's understanding of the topic, with a quoted example from the body of the IA. End with the 3 highest-leverage changes for the redraft."

Internal links to IB History IA workflow:

• → /blog/ai-tutor-ib-2026 — post-33, IB DP parent guide
• → /blog/ai-tutor-ib-extended-essay-playbook-2026 — post-62, the EE workflow that mirrors History IA source-evaluation discipline

---

IB English A Individual Oral: 15 minutes, externally assessed, the most underscripted IA

The English A Individual Oral is a 15-minute oral presentation on a globally-minded literary work paired with a work studied in class. It is graded on four criteria totalling 25 marks: Knowledge, Understanding, and Interpretation (0–10), Analysis and Evaluation (0–10), Focus, Organisation, and Development (0–5). The grade 7 oral scores 22–25/25.

The single biggest grade C killer in English A IO: the candidate summarises the plot of both works in the first 5 minutes and then has only 10 minutes left for the analysis. The grade 7 oral opens with the thesis sentence ("The treatment of power in [work 1] and [work 2] is structurally similar but politically opposed — both texts use silence as a metaphor for dissent, but the silence serves a different ideological function in each"), supports that thesis with two short textual extracts, and ends with a globally-minded reflection.

AI tutor prompt for the English A Individual Oral:

"Read this 15-minute English A Individual Oral transcript. For each criterion (Knowledge/Understanding/Interpretation, Analysis/Evaluation, Focus/Organisation/Development) give a mark out of the available marks. Specifically flag: (a) the percentage of the oral spent summarising plot rather than analysing — if more than 25% of the oral is summary, the candidate will cap at grade 5; (b) whether the globally-minded discussion links both works to a real-world context beyond the texts themselves; (c) whether the Analysis criterion addresses the candidate's own interpretation of how a literary feature functions in the text, rather than restating what a critic has said. End with the 3 highest-leverage changes the candidate should make."

---

IB Maths AA / AI Internal Assessment: the exploration that is half written work, half mathematical reasoning

The Maths IA is a 6–12 page exploration of a mathematical topic the student finds personally engaging. It is graded on five criteria totalling 20 marks: Presentation (0–4), Mathematical Communication (0–4), Personal Engagement (0–4), Reflection (0–4), Use of Mathematics (0–4).

The single biggest grade C killer in Maths IA: the student does the mathematics correctly but does not communicate it to a reader who is not the student. A grade 5 exploration reads like a calculation log. A grade 7 exploration reads like a 12-page mathematical essay, with clear section headings, full sentences explaining what each equation represents, definitions of every variable, and at least 2 diagrams.

AI tutor prompt for the Maths IA:

"Mark this Maths IA against the IB Maths IA rubric. For each of the five criteria (Presentation, Mathematical Communication, Personal Engagement, Reflection, Use of Mathematics) give a mark out of the available marks. Specifically flag: (a) whether the Use of Mathematics criterion is met at the SL or HL level (the rubric requires 'mathematics that is at the same level or slightly beyond the level of the course'), (b) whether the Personal Engagement criterion is demonstrated through the student's own examples, own data, or own phrasing of a known result (not through statements like 'I found this interesting'), (c) whether the Reflection section identifies a specific mathematical insight that emerged from the exploration, with a worked example showing how the insight was discovered. End with the 3 highest-leverage changes."

Internal links to IB Maths IA workflow:

• → /blog/ai-tutor-ib-extended-essay-playbook-2026 — post-62, EE workflow that mirrors the Maths IA research methodology
• → /blog/ai-tutor-ks3-year-9-options-ib-myp-2026 — post-35, KS3 MYP to IB Maths transition

---

IB Psychology IA: the replication study that mirrors an actual experiment

The Psychology IA is a 6–12 page replication of a published psychological study, conducted on a small sample (typically 10–30 participants). It is graded on four criteria totalling 22 marks: Research Design (0–6), Data Analysis and Interpretation (0–6), Discussion and Evaluation (0–6), Communication (0–4).

The single biggest grade C killer in Psychology IA: the Discussion section lists the limitations of the study without linking them to the actual results. A grade 5 Discussion reads "The sample size was small and the participants were all from one school, so the results may not generalise." A grade 7 Discussion reads "The sample size of 18 participants was sufficient to detect a medium effect size (Cohen's d = 0.65) but underpowered to detect a small effect (d = 0.3). The single-school sampling limits generalisability to adolescents in urban UK schools; a multi-school replication would be needed to establish whether the effect holds across socioeconomic contexts. The use of a self-report measure introduces social desirability bias, which likely inflated the mean score on the conformity scale by approximately 0.4 standard deviations."

AI tutor prompt for the Psychology IA:

"Mark this Psychology IA against the IB Psychology IA rubric. For each of the four criteria (Research Design, Data Analysis and Interpretation, Discussion and Evaluation, Communication) give a mark out of the available marks. Specifically flag: (a) whether the Discussion section links each named limitation to a specific quantitative consequence on the results (not just 'small sample' but 'sample of N = 18 underpowered to detect effect sizes smaller than d = 0.6'), (b) whether the Data Analysis section uses inferential statistics appropriate for the research design (t-test for between-subjects, paired t-test for within-subjects, chi-square for categorical), (c) whether the Communication criterion is met through APA-style referencing and structured headings. End with the 3 highest-leverage changes."

---

IB Economics IA: the 6-page commentary on a real-world article

The Economics IA is a 6-page commentary on a published article that relates to the IB Economics syllabus. It is graded on five criteria totalling 25 marks: Real World Issue (0–5), Economic Theory (0–5), Use of Diagrams (0–5), Analysis and Evaluation (0–5), Structure (0–5).

The single biggest grade C killer in Economics IA: the commentary explains the article rather than applying economic theory to it. A grade 5 IA reads "The article says that inflation rose in country X. This is because of cost-push factors." A grade 7 IA reads "The article documents a 4.7% CPI increase in country X between Q1 and Q3 2025. Applying the AD-AS model, this is consistent with a positive supply shock given the observed unemployment trajectory; the alternative demand-pull explanation is less consistent with the falling investment data in Table 2 of the article. The diagram below shows the AD-AS shift with the SRAS curve moving leftward by 4.7%."

AI tutor prompt for the Economics IA:

"Mark this Economics IA against the IB Economics IA rubric. For each of the five criteria (Real World Issue, Economic Theory, Use of Diagrams, Analysis and Evaluation, Structure) give a mark out of the available marks. Specifically flag: (a) whether the Real World Issue is from the past 12 months and is a recognised macroeconomic or microeconomic event, (b) whether at least 3 economic theory concepts from the IB syllabus are applied explicitly to the article (not just mentioned), (c) whether at least 2 original diagrams are included, drawn by the student (not copied from a textbook), with full labels. End with the 3 highest-leverage changes."

Internal links to IB Economics IA workflow:

• → /blog/ai-tutor-ib-2026 — post-33, IB DP parent guide

---

IB Geography IA: the fieldwork report with data-collection specificity

The Geography IA is a 2,500-word fieldwork report based on primary data the student collects at a site visit. It is graded on five criteria totalling 25 marks: Fieldwork Question and Geographic Context (0–5), Methods of Investigation (0–5), Quality and Treatment of Data (0–5), Analysis and Interpretation (0–5), Conclusion and Evaluation (0–5).

The single biggest grade C killer in Geography IA: the data collection methodology is described in general terms rather than with site-specific details. A grade 5 IA reads "We collected data at 10 sites." A grade 7 IA reads "We collected data at 10 sites along a 1.4 km transect of the River Mole, spaced at 150 m intervals to capture the downstream gradient in sediment size and channel width. At each site we measured channel width with a 30 m tape at 3 cross-sections, depth with a metre rule at 5 points across each cross-section, and velocity using a flow probe at 60% depth for 60 seconds. The transect was completed between 09:00 and 13:00 on 14 October 2026 to minimise temporal variability in discharge."

AI tutor prompt for the Geography IA:

"Mark this Geography IA against the IB Geography IA rubric. For each of the five criteria (Fieldwork Question and Geographic Context, Methods of Investigation, Quality and Treatment of Data, Analysis and Interpretation, Conclusion and Evaluation) give a mark out of the available marks. Specifically flag: (a) whether the Methods of Investigation criterion specifies where, when, and how each piece of data was collected with sufficient detail for the study to be replicated, (b) whether the Quality and Treatment of Data section includes at least one statistical technique (mean, median, standard deviation, chi-square, Spearman's rank), (c) whether the Conclusion is linked back to the original fieldwork question with a 1-sentence yes/no/qualified answer before the explanation. End with the 3 highest-leverage changes."

---

The 4 highest-leverage IA habits that lift the grade across all subjects

The per-subject AI tutor prompts above give subject-specific marking. But the 4 habits below lift the grade across every IA subject because they target the criteria that share weight in every rubric:

Habit 1 — Write the Evaluation / Reflection section last, not first. Most students write the IA in the order it appears: introduction, methodology, results, discussion, conclusion. Grade 7 students write the IA in the order the rubric demands: research question, methodology, data, and then go back and write the Evaluation / Reflection section last with the full data in front of them. This is the single highest-leverage writing habit because it forces the Evaluation to be data-grounded rather than hypothesis-grounded.

Habit 2 — Quote the rubric into the AI tutor prompt. When you mark an IA draft with an AI tutor, paste the actual IB rubric text (not a paraphrase) into the prompt. The AI tutor's marking is only as accurate as the rubric it has been given. The IB publishes the rubric for every subject in the subject guide, free to download from the IB programme resource centre.

Habit 3 — Do 3 redrafts, not 1. The IA is a 6-month project. The teacher marks one draft in late February or early March. The student then has 3–4 weeks to redraft. The grade 7 student does 3 redrafts: a structural redraft (does each section answer the rubric criterion?), an argument redraft (does each paragraph argue rather than describe?), and a polish redraft (figures, bibliography, word count). Each redraft is 4–5 days of focused work, not 4–5 hours.

Habit 4 — Submit the IA to the IB's MyIB subject guide exemplar list. The IB publishes exemplar IAs for every subject at every grade level (1–7). Reading 3 grade 7 exemplars from the same subject teaches the student more about the rubric than any other resource. The exemplars are free for IB-coordinator-registered schools and available to parents from the IB store.

---

How an AI tutor should be prompt-engineered to mark an IB IA

A general-purpose AI tutor is not the same as an IB-aware AI tutor. The IB-aware tutor needs three prompt layers:

Layer 1 — The rubric layer. The system prompt should include the full IB IA rubric for the subject, the 2024-onwards moderation emphasis, and the grade-band descriptors for marks 1–2, 3–4, 5–6, and 7. Without this layer, the AI tutor marks against generic essay criteria and the student's grade lifts by 1 IB point at best.

Layer 2 — The exemplar layer. The system prompt should reference 2–3 IB-published grade 7 exemplars and 1 grade 5 exemplar for the subject. This grounds the AI tutor's marking in the IB's own quality bar rather than a generic academic standard.

Layer 3 — The student-history layer. The system prompt should know what subject the student is studying, what IA topic they have chosen, what draft number this is, and what marks they received on the previous draft (if any). This lets the AI tutor track grade lift across drafts and identify any criterion that has stalled at the same mark for two consecutive drafts.

An AI tutor with all three layers is what the rest of this playbook calls a "rubric-aware AI tutor". An AI tutor without these layers is a generic essay-marking tool that misses the per-subject IB specificity.

---

The teacher + AI tutor workflow: who marks what

The AI tutor does not replace the teacher. The AI tutor does the rubric-aware marking pass in February, the teacher does the rubric-aware marking pass in March, and the student has a 3–4 week window to apply both sets of feedback in the final redraft.

The specific roles:

• Teacher (March): marks the IA against the IB rubric, gives a provisional mark out of the available total, writes 3–5 paragraphs of feedback, returns the work to the student, then meets the student for a 20-minute feedback conference. This is the official IB-required moderation step.
• AI tutor (February): marks the IA draft against the IB rubric, gives a mark and 3 highest-leverage changes, does not meet the student. This is the unofficial pre-emptive step that ensures the teacher sees a draft that has already addressed the obvious rubric gaps.
• Student (between February and March): applies the AI tutor feedback, does 1–2 redrafts, submits to the teacher for the official March marking pass.
• Student (between March and April): applies the teacher feedback, does the final redraft, submits for IB upload.

The teacher–AI tutor split works because the teacher's time is finite (a class of 25 IB Biology students means 25 IA second drafts to mark in March, which is 50+ hours of marking) and the AI tutor's time is unbounded. The AI tutor does the work that does not require a teacher–student relationship; the teacher does the work that does.

---

The single biggest mistake IB students make on the IA

After working with hundreds of IB students across 7 IA subjects, the single most common grade C mistake is: the student writes an IA that the teacher would mark at grade 5 in February, then does not redraft it. They submit the grade-5 draft to the teacher in March, receive the grade 5 mark, and submit it to the IB in April. They never lift the grade because they did not apply the AI tutor feedback before the teacher mark.

The single highest-leverage change a student can make to their IB IA grade is: mark the draft with an AI tutor in February, redraft based on the AI tutor's feedback in early March, submit the redraft to the teacher for the March marking pass. This single workflow change lifts the median IA grade from 5 to 6 — and a 6 across 6 IAs is the difference between a 36-point diploma and a 42-point diploma.

---

The 6-month IA timeline (May 2027 session)

Month	Stage	Owner	AI tutor role
Sep 2026	Topic and research question	Student	Stress-test research question against scope criterion
Oct 2026	Source / methodology / pilot	Student	Flag methodology risks (confounding variables, source selection)
Nov 2026	Data collection / reading	Student	None — student is doing the work
Dec 2026	Drafting — introduction + methodology	Student	None — student is doing the writing
Jan 2027	First full draft complete	Student	First rubric-aware mark
Feb 2027	Redraft based on AI tutor feedback	Student	Re-mark the redraft against the same rubric
Mar 2027	Teacher-marked second draft + 20-min conference	Teacher + Student	Optional third mark if needed
Apr 2027	Final submission to IB	Teacher	Format and citation check

The dates assume a May 2027 exam session with IA submission in April 2027. For the November 2027 session the timeline shifts by 6 months.

---

Sources and references

This playbook cites the following IB and IB-adjacent sources:

1. IB Diploma Programme Subject Guides 2024 — Biology, Chemistry, Physics, History, English A, Mathematics: Analysis and Approaches, Mathematics: Applications and Interpretation, Psychology, Economics, Geography. Free download from the IB programme resource centre.
2. IB Subject Reports 2024 and 2025 — published annually by the IB after each exam session. Contains examiner commentary on the most common grade-C errors per subject per IA.
3. IB Teacher Support Materials 2024 — internal materials for IB-registered teachers that include worked grade 7 and grade 5 exemplar IAs.
4. IB Extended Essay Guide 2024 — covers the EE assessment criteria that mirror the IA criteria in architecture.
5. IB Coordinator Notes 2024 — moderation procedures, including the 5-sample-per-subject-per-school moderation spot check.
6. IBO Statistical Bulletin 2024 — global IA grade distributions and moderation outcomes.
7. Internal Grademy data 2025–2026 — IA grade lift observed in cohort of 12 IB Diploma students using the rubric-aware AI tutor workflow above.

---

Closing — what to do this week

If you are an IB Diploma student entering the May 2027 exam session, the single most valuable thing you can do this week is pick the IA topic for the subject you find hardest, draft a research question, paste it into an AI tutor with the rubric-aware prompt from this playbook, and read the response. The response will tell you whether the research question is too broad, too narrow, or appropriately scoped. You will have an answer in 20 minutes that you would otherwise have waited until February to discover.

If you are the parent of an IB Diploma student, the single most valuable thing you can do this week is sit with your child for 20 minutes and ask them to walk you through the IA research question for one subject. If they cannot explain the question to a non-expert in 3 sentences, the question is not yet clear enough to write a grade 7 IA against.

If you are an IB IA subject teacher, the single most valuable thing you can do this week is paste the rubric for your subject into your AI tutor of choice and run it against 3 grade 7 exemplars and 3 grade 5 exemplars from your subject guide. The AI tutor will then have the rubric grounding it needs to give your students accurate pre-moderation feedback in February, which is the highest-leverage week of the IA timeline.

Grademy's IB IA workflow is built around the rubric-aware AI tutor prompts above. The system marks a student's IA draft against the per-subject IB rubric, identifies the 3 highest-leverage changes, and tracks the grade lift across redrafts — exactly the workflow a class of 25 IB Biology students needs in February 2027.

• IB Biology HL AI tutor playbook 2026 — for IB Biology HL candidates writing a Bio IA: the IA is the highest-leverage 20 percent of the final grade because it is internally marked + externally moderated and scored against the 5 IB criteria (Personal Engagement + Exploration + Analysis + Evaluation + Communication); the Bio HL IA workflow benefits from the IB IA general rubric + the Bio HL-specific experimental-design depth (Daphnia heart-rate as a pharmacology model, enzyme-kinetics at varying substrate concentrations, plant-transpiration at varying light intensity, yeast fermentation rates at varying temperatures)

• IB Biology HL AI tutor playbook 2026 — for IB Biology HL candidates writing a Bio IA: the IA is the highest-leverage 20 percent of the final grade because it is internally marked + externally moderated and scored against the 5 IB criteria (Personal Engagement + Exploration + Analysis + Evaluation + Communication); the Bio HL IA workflow benefits from the IB IA general rubric + the Bio HL-specific experimental-design depth (Daphnia heart-rate as a pharmacology model, enzyme-kinetics at varying substrate concentrations, plant-transpiration at varying light intensity, yeast fermentation rates at varying temperatures)

• IB Chemistry HL AI tutor playbook 2026 — for IB Chemistry HL candidates writing a Chem IA: the IA is the highest-leverage 20 percent of the final grade because it is internally marked + externally moderated and scored against the 5 IB criteria (Personal Engagement + Exploration + Analysis + Evaluation + Communication); the Chem HL IA workflow benefits from the IB IA general rubric + the Chem HL-specific experimental-design depth (rate-of-reaction at varying temperatures using iodine-clock, aspirin-synthesis percent-yield + melting-point purity, electroplating-copper mass-vs-time Faraday-law analysis, vitamin-C titration by iodometric-back-titration)

• IB Chemistry HL AI tutor playbook 2026 — for IB Chemistry HL candidates writing a Chem IA: the IA is the highest-leverage 20 percent of the final grade because it is internally marked + externally moderated and scored against the 5 IB criteria (Personal Engagement + Exploration + Analysis + Evaluation + Communication); the Chem HL IA workflow benefits from the IB IA general rubric + the Chem HL-specific experimental-design depth (rate-of-reaction at varying temperatures using iodine-clock, aspirin-synthesis percent-yield + melting-point purity, electroplating-copper mass-vs-time Faraday-law analysis, vitamin-C titration by iodometric-back-titration)

Related: the IB English A Individual Oral as the IB Internal Assessment

For IB English A: Language and Literature, the IB Internal Assessment is the Individual Oral (10 minutes preparation + 10 minutes delivery + 5 minutes Q&A, 15 minutes of recorded delivery, plus 5 minutes of Q&A, totalling 20 minutes, worth 30 percent for SL or 20 percent for HL). The Individual Oral is structured differently from IB Biology HL + IB Chemistry HL + IB Math AA + IB Physics HL IAs — it is a 10-minute oral analysis of a global issue extracted from one Part 3 work paired with one Part 4 work (HL only) or with one Part 3 work (SL only). See our <a href="/blog/ai-tutor-ib-english-a-language-literature-playbook-2026" class="text-[var(--brand-coral)] underline-offset-4 hover:underline font-semibold">AI tutor for IB English A: Language and Literature 2026</a> for the Individual Oral + Paper 1 + Paper 2 workflow.

Pairs with IB Physics HL (post-157)

The IB Physics HL playbook (post-157) + [this playbook] form the IB Diploma + IB Physics Higher Level + IBO-mastery + Year-13-physics + first-year-undergraduate-Physics + Engineering + Medicine + Architecture + Astrophysics pair for the IB-Diploma-Programme Year-12 + Year-13 candidates + IB-Bilingual + IB-Course-Companion + Pre-Engineering + Pre-Medicine + Pre-Architecture + Pre-Physics + Pre-Astrophysics cohorts at IB-World-Schools in Switzerland + Singapore + Hong-Kong + UAE + UK + US + Canada + Australia + New-Zealand + Japan + Korea + China + India + Pakistan + Bangladesh + Indonesia + Malaysia + Thailand + Vietnam + Philippines + Egypt + South-Africa + Kenya + Nigeria + Ghana + Saudi + Qatar + Bahrain + Kuwait + Oman + Jordan + Lebanon + Brazil + Mexico + Argentina + Chile + Colombia + Peru + Russia + Turkey + Iran + Iraq + Israel targeting MIT + Imperial + Cambridge + Oxford + ETH-Zurich + EPFL + NUS + NTU + HKUST + KAIST + IIT + IISc + UCL + LSE + Warwick + Edinburgh + Manchester + Bristol + Durham + KCL admissions. The IB Physics HL 8-core-topics (Topic 1 Measurements + Uncertainties + Topic 2 Mechanics + Topic 3 Thermal Physics + Topic 4 Waves + Topic 5 Electricity + Magnetism + Topic 6 Circular Motion + Gravitation + Topic 7 Atomic + Nuclear + Topic 8 Energy Production) + 4-additional-higher-level-topics (Topic 9 Wave Phenomena-HL + Topic 10 Fields-HL + Topic 11 Electromagnetic Induction-HL + Topic 12 Quantum + Nuclear Physics-HL) + 12-practical-lab-groups + Internal-Assessment-Individual-Investigation + Paper-1A-MCQ + Paper-1B-data-based + Paper-2-extended-response + Paper-3-Optional-Topic + calculus-based + critical-thinking-driven + extended-response-required + IA-scientific-writing-assessed + 7-IA-Individual-Investigation-competencies + 4-Paper-1A-MCQ-competencies + 4-Paper-1B-data-based-competencies + 4-Paper-2-extended-response-competencies + 4-Paper-3-Optional-Topic-competencies + 8-Topic-mastery-competencies + AI-tutor-prompt-library + score-7-rubric covers everything the IB-World-School-cohort needs to bridge Year-12 + Year-13-physics + IB-Diploma + IBO-mastery-track + first-year-undergraduate-Physics + Engineering + Medicine + Architecture + Astrophysics-university-credit + STEM-Admissions-readiness at MIT + Imperial + Cambridge + Oxford + ETH-Zurich + EPFL + NUS + NTU + HKUST + KAIST + IIT + IISc + UCL + LSE + Warwick + Edinburgh + Manchester + Bristol + Durham + KCL. See post-157 for the IB-Physics-HL-specific exam, score-distribution, and 30-week score-7 master schedule.

Pairs with IB Math HL Analysis & Approaches (post-158)

The IB Math HL Analysis & Approaches playbook (post-158) + [this playbook] form the IB Diploma + IB Mathematics Higher Level + IBO-mastery + Year-13-mathematics + first-year-undergraduate-Mathematics + Engineering + Physics + Computer-Science + Medicine + Architecture + Economics + Finance + Statistics pair for the IB-Diploma-Programme Year-12 + Year-13 candidates + IB-Bilingual + IB-Course-Companion + Pre-Engineering + Pre-Physics + Pre-Computer-Science + Pre-Mathematics + Pre-Medicine + Pre-Architecture + Pre-Economics + Pre-Finance + Pre-Statistics cohorts at IB-World-Schools in Switzerland + Singapore + Hong-Kong + UAE + UK + US + Canada + Australia + New-Zealand + Japan + Korea + China + India + Pakistan + Bangladesh + Indonesia + Malaysia + Thailand + Vietnam + Philippines + Egypt + South-Africa + Kenya + Nigeria + Ghana + Saudi + Qatar + Bahrain + Kuwait + Oman + Jordan + Lebanon + Brazil + Mexico + Argentina + Chile + Colombia + Peru + Russia + Turkey + Iran + Iraq + Israel targeting MIT + Imperial + Cambridge + Oxford + ETH-Zurich + EPFL + NUS + NTU + HKUST + KAIST + IIT + IISc + UCL + LSE + Warwick + Edinburgh + Manchester + Bristol + Durham + KCL admissions. The IB Math HL AA 5-core-topics (Topic 1 Number and Algebra + Topic 2 Functions + Topic 3 Geometry and Trigonometry + Topic 4 Statistics and Probability + Topic 5 Calculus) + 5-HL-additional-topics (Algebra + Logarithmic + Exponential + Functions + Trigonometric + Functions + Vectors + Statistics + Probability) + Internal-Assessment-Mathematical-Exploration + Paper 1 + Paper 2 + Paper 3 Optional-Topic + proof-based + calculus-based + extended-response-required + IA-mathematical-exploration-assessed + 5-proof-by-induction-competencies + 6-differential-equations-competencies + 4-linear-algebra-competencies + 5-complex-numbers-competencies + 4-vectors-in-3D-competencies + 5-calculus-advanced-series-competencies + 7-prompt-AI-tutor-library + score-7-rubric covers everything the IB-World-School-cohort needs to bridge Year-12 + Year-13-mathematics + IB-Diploma + IBO-mastery-track + first-year-undergraduate-Mathematics + Engineering + Physics + Computer-Science + Medicine + Architecture + Economics + Finance + Statistics-university-credit + STEM-Admissions-readiness at MIT + Imperial + Cambridge + Oxford + ETH-Zurich + EPFL + NUS + NTU + HKUST + KAIST + IIT + IISc + UCL + LSE + Warwick + Edinburgh + Manchester + Bristol + Durham + KCL. See post-158 for the IB-Math-HL-AA-specific exam, score-distribution, and 30-week score-7 master schedule.

Pairs with IB Math HL Applications & Interpretation (post-159)

The IB Math HL Applications & Interpretation playbook (post-159) + [this playbook] form the IB Diploma + IB Mathematics Higher Level + IBO-applied-track + Year-13-applied-mathematics + first-year-undergraduate-Economics + Finance + Business + Statistics + Actuarial-Science + Psychology + Biology + Environmental-Science + Data-Science + Computer-Science pair for the IB-Diploma-Programme Year-12 + Year-13 candidates + IB-Bilingual + IB-Course-Companion + Pre-Economics + Pre-Finance + Pre-Business + Pre-Psychology + Pre-Biology + Pre-Environmental-Science + Pre-Statistics + Pre-Data-Science + Pre-Computer-Science cohorts at IB-World-Schools in Switzerland + Singapore + Hong-Kong + UAE + UK + US + Canada + Australia + New-Zealand + Japan + Korea + China + India + Pakistan + Bangladesh + Indonesia + Malaysia + Thailand + Vietnam + Philippines + Egypt + South-Africa + Kenya + Nigeria + Ghana + Saudi + Qatar + Bahrain + Kuwait + Oman + Jordan + Lebanon + Brazil + Mexico + Argentina + Chile + Colombia + Peru + Russia + Turkey + Iran + Iraq + Israel targeting MIT + Imperial + Cambridge + Oxford + LSE + Warwick + UCL + LBS + Bocconi + HEC-Paris + ESSEC + INSEAD + Sciences-Po + NUS + NTU + HKUST + KAIST + UCL + Edinburgh + Manchester + Bristol + Durham + KCL admissions. The IB Math HL AI 5-core-topics (Topic 1 Number and Algebra + Topic 2 Functions + Topic 3 Geometry and Trigonometry + Topic 4 Statistics and Probability + Topic 5 Calculus) + 5-HL-additional-topics (Algebra + Logarithmic + Exponential + Functions + Trigonometric + Functions + Vectors + Statistics + Probability) + Internal-Assessment-Modelling-Exploration + Paper 1 + Paper 2 + Paper 3 Optional-Topic + modelling-driven + applied-calculus-required + statistics-heavy + GDC-required + IA-modelling-exploration-assessed + 5-modelling-and-optimization-competencies + 6-statistics-and-probability-applied-competencies + 4-financial-modelling-competencies + 5-discrete-modelling-competencies + 4-optimisation-applications-competencies + 5-calculus-applied-series-competencies + 7-prompt-AI-tutor-library + score-7-rubric covers everything the IB-World-School-cohort needs to bridge Year-12 + Year-13-applied-mathematics + IB-Diploma + IBO-applied-mastery-track + first-year-undergraduate-Economics + Finance + Business + Statistics + Actuarial-Science + Psychology + Biology + Environmental-Science + Data-Science + Computer-Science-university-credit + Applied-Admissions-readiness at MIT + Imperial + Cambridge + Oxford + LSE + Warwick + UCL + LBS + Bocconi + HEC-Paris + ESSEC + INSEAD + Sciences-Po + NUS + NTU + HKUST + KAIST + UCL + Edinburgh + Manchester + Bristol + Durham + KCL. See post-159 for the IB-Math-HL-AI-specific exam, score-distribution, and 30-week score-7 master schedule.

Pairs with IB Chemistry HL (post-160)

The IB Chemistry HL playbook (post-160) + [this playbook] form the IB Diploma + IB Higher-Level + IBO-applied-track + Year-13-chemistry + first-year-undergraduate-Medicine + Pharmacy + Biology + Chemical-Engineering + Biochemistry + Dentistry + Veterinary-Science pair for the IB-Diploma-Programme Year-12 + Year-13 candidates + IB-Bilingual + IB-Course-Companion + Pre-Medicine + Pre-Pharmacy + Pre-Biology + Pre-Chemical-Engineering + Pre-Biochemistry + Pre-Dentistry + Pre-Veterinary cohorts at IB-World-Schools in Switzerland + Singapore + Hong-Kong + UAE + UK + US + Canada + Australia + New-Zealand + Japan + Korea + China + India + Pakistan + Bangladesh + Indonesia + Malaysia + Thailand + Vietnam + Philippines + Egypt + South-Africa + Kenya + Nigeria + Ghana + Saudi + Qatar + Bahrain + Kuwait + Oman + Jordan + Lebanon + Brazil + Mexico + Argentina + Chile + Colombia + Peru + Russia + Turkey + Iran + Iraq + Israel targeting MIT + Imperial + Cambridge + Oxford + ETH-Zurich + EPFL + NUS + NTU + HKUST + KAIST + UCL + Edinburgh + Manchester + Bristol + Durham + KCL admissions. The IB Chemistry HL 11-core-topics (Topic 1 Stoichiometric Relationships + Topic 2 Atomic Structure + Topic 3 Periodicity + Topic 4 Chemical Bonding + Topic 5 Energetics + Topic 6 Chemical Kinetics + Topic 7 Chemical Equilibrium + Topic 8 Acids and Bases + Topic 9 Redox Processes + Topic 10 Organic Chemistry + Topic 11 Measurement and Data Processing) + 4-additional-higher-level-topics (advanced-stoichiometry + advanced-energetics + advanced-kinetics + advanced-equilibrium + advanced-acids-bases + advanced-redox + advanced-organic-mechanism + advanced-spectroscopy-NMR-IR-MS-UV + transition-element-complexes + ligand-field-theory + chirality-and-stereochemistry + organic-synthesis-strategies + multi-step-synthesis) + Internal-Assessment-Individual-Investigation + Paper 1A MCQ + Paper 1B data-based + Paper 2 extended-response + Paper 3 Optional-Topic-Biochemistry-Energy-Medicinal-Chemistry-Materials-Science + stoichiometry-driven + energetics-required + organic-mechanism-heavy + kinetics-required + IA-individual-investigation-assessed + 5-stoichiometry-competencies + 5-atomic-structure-competencies + 4-periodicity-competencies + 5-bonding-competencies + 5-energetics-competencies + 5-kinetics-competencies + 5-equilibrium-competencies + 6-acids-bases-competencies + 5-redox-competencies + 8-organic-chemistry-competencies + 5-measurement-analysis-competencies + 7-AHL-advanced-spectroscopy-competencies + 7-AHL-advanced-mechanism-competencies + 7-prompt-AI-tutor-library + score-7-rubric covers everything the IB-World-School-cohort needs to bridge Year-12 + Year-13-chemistry + IB-Diploma + IBO-applied-mastery-track + first-year-undergraduate-Chemistry + Medicine + Pharmacy + Biology + Chemical-Engineering + Biochemistry + Dentistry + Veterinary-Science-university-credit + Applied-Admissions-readiness at MIT + Imperial + Cambridge + Oxford + ETH-Zurich + EPFL + NUS + NTU + HKUST + KAIST + UCL + Edinburgh + Manchester + Bristol + Durham + KCL. See post-160 for the IB-Chemistry-HL-specific exam, score-distribution, and 30-week score-7 master schedule.