The AI landscape doesn't move in one direction — it lurches. Some techniques leap from experiment to table stakes in a single quarter; others stall against regulatory walls, technical ceilings, or organisational inertia that no amount of hype can dislodge. Knowing which is which is the hard part. The State of Play cuts through the noise with a rigorously maintained index of AI techniques across every major business domain — classified by maturity, evidenced by real-world adoption, and updated daily so you always know where you stand relative to the field. Stop guessing. Start knowing.
A daily newsletter distilling the past two weeks of movement in a domain or two — delivered to your inbox while the index updates in the background.
Each dot marks the weighted maturity of practices within a domain — hover for a brief summary, click for more detail
AI that designs curricula, generates learning paths, creates course content, and produces lesson plans aligned to learning objectives. Includes standards-aligned content creation and prerequisite mapping; distinct from question generation which creates assessment items rather than instructional content.
AI-powered curriculum design has matured into a leading-edge but structurally constrained practice. By June 2026, the evidence is unambiguous: deployment at scale is real, time savings are quantified, yet learning outcomes depend entirely on how curricula are pedagogically designed—not on tool sophistication. UK-wide evidence (May 2026) documents 54% of teachers use AI for lesson planning, 50% for quizzes, 45% for test materials, with EEF/NFER controlled trials proving 31% reduction in prep time (~25 min/lesson) with no quality loss by blind expert review. US adoption mirrors scale (68-72% weekly use), and Stanford's analysis of 150,000+ teacher prompts confirms curriculum design is the dominant substantive AI use case in education. Yet the learning-outcomes picture is sobering. The OECD Digital Education Outlook 2026 establishes a critical threshold: "GenAI without clear teaching principles enhances performance with no real learning gains." A meta-analysis of 22 studies (2022–2025) shows positive effects only when pedagogical intent is explicit (g=0.572), while 90% of generic AI lesson materials constrain thinking to recall levels. Large-scale institutional deployments (Tec de Monterrey with 64K+ students, Copenhagen Business Academy with full course integration, JPMorgan's 300K-person curriculum) show the practice is production-ready. But they also reveal the hard requirement: every AI-generated lesson plan requires expert human review and pedagogical validation before deployment. The defining tension persists unchanged: mainstream adoption and time savings coexist with universal "supervision debt" (mandatory review), training gaps (29% of teachers lack formal AI instruction), governance deficits (87% of schools have no formal AI policy), and structural barriers preventing classroom-ready outputs without expert curation. The practice has crossed into leading-edge through deployment maturity and institutional infrastructure—not through solving the underlying design problem.
Adoption has reached true mainstream scale with quantified productivity gains and emerging agentic-system deployment. June 2026 evidence shows: UK nationally representative survey (8,000-10,000 teachers) documents 54% use AI for lesson planning, 50% for quizzes, 45% for test materials, 39% for school reports; EEF/NFER controlled trial (68 schools, 259 science teachers) measured ChatGPT-assisted lesson planning reducing prep time to 31% of control baseline (~25 minutes per lesson) with blind expert review finding no pedagogical quality loss. US adoption mirrors this pattern: 68-72% of K-12 teachers use AI weekly; Stanford's analysis of 150,000+ teacher prompts confirms ~50% relate to curriculum design (standards unpacking, example generation, materials alignment, accessibility adaptation). The vendor ecosystem is consolidating and innovating: MagicSchool operates at $44.9M ARR across 13,000+ schools and 6M+ users; Khanmigo reaches 700k users across 380+ districts with Harvard/Stanford RCT validation; next-generation tools (Atomic Jolt, PepperMill, Kuraplan) automate standards alignment and gap analysis; emerging agentic approaches (27% of L&D organizations active, 39% cautious but interested) shift from reactive drafting to proactive multi-step workflows. Policy institutionalization accelerates globally: 134 US state bills, Georgia/Mississippi mandate AI in graduation standards; India launches "AI Literacy for Teachers" (targeting 1M teachers by 2027); EU and UK government initiatives (DfE/DSIT) develop curriculum-aligned safety standards and content infrastructure. Large-scale institutional deployments (Tec de Monterrey: 64K+ students with 20–48 point score gains; Copenhagen Business Academy: no-code RAG for course-specific assistants; JPMorgan: 300K-person "AI Made Easy" curriculum) demonstrate production-stage maturity.
Yet adoption vastly outpaces implementation capacity. Only 29% of US teachers received formal AI training (May 2026); 87% of schools operate without formal AI governance. Quality outcomes remain inconsistent: practitioner reviews find basic worksheets acceptable (71%) but assessment items requiring higher-order thinking only 38% usable, IEP recommendations 36% good. The OECD establishes a design threshold: generic tools cause "cognitive offloading" with 17% worse exam performance despite higher practice scores—validating critical signal that curriculum design must prioritize "slow AI" (iterative, scaffolded) over "fast AI" (fluent one-shot generation). Critical research documents that sequential project design with high element interactivity outperforms open-ended approaches when AI is present. The universal "supervision debt" persists—every output requires expert validation. Field-building is underway (Digital Promise/TNTP partnership, $23M National Academy for AI Instruction targeting 400k teachers, NSF AmplifyGAIN research centers) but faces systemic barriers: insufficient PD infrastructure, weak learning science grounding in most tools, and widespread institutional inertia. Critical assessment from 50+ stakeholder interviews documents tools functioning as "sophisticated photocopiers" rather than pedagogical partners. The result: 68% adoption coexists with structural constraints—training gaps, governance deficits, and lack of pedagogical frameworks—that prevent scaling beyond early-adopter contexts. Emerging signals show the field is shifting from generative drafting to agentic systems and pedagogically-grounded design, but maturity remains concentrated in institutions with dedicated design capacity.
— Identifies fundamental curriculum design gaps for AI era: unit of work shifted from typing to specifying; six core AI-native competencies missing from traditional curricula; frames 24-month window for AI-native training advantage.
— 2026 comprehensive tool comparison documenting market maturity: standards alignment, adaptive sequencing, auto-generated assessments, bias guardrails; validates practice as institutional/scaled deployment across K-12, higher ed, and corporate contexts.
— Empirical evidence of critical limitation: AI-assisted work improves output quality (M=7.62) without strengthening actual learning mastery (M=5.55), producing average AI-Learning Gap of 2.07; signals fundamental effectiveness constraint for AI-generated curriculum.
— Domain-specific curriculum framework with identified barriers (88% student recognition of AI importance but minimal formal instruction, 92% trainee inadequacy); addresses educator readiness and infrastructure gaps limiting integration.
— Deep critical assessment of AI's fundamental limitations: context blindness, lack of empathy, bias propagation; proposes hybrid human-AI model as maturity path; documents adoption barriers rooted in capability constraints.
— India's CBSE mandatory Computational Thinking & AI curriculum redesign (Classes 3-8: 50-100 hrs/yr) affecting millions of students; composite skill labs required by Aug 2027; documents institutional curriculum design at national scale.
— Italy national curriculum guidelines (effective Sept 2026) mandate AI-supported 6-step pedagogical workflow; emphasizes skill-based pathways and teacher-led instructional leadership; represents policy-level adoption of AI-supported curriculum generation.
— Direct evidence of educator upskilling for curriculum design: 60+ educators trained on AI instructional design, assessment creation, differentiation using Gemini and NotebookLM; addresses rural equity gap in access to AI curriculum training.