The State of SCORM 2026
By the SCORM Converter Team · Updated June 22, 2026
Based on telemetry from 107 real SCORM packages processed in our regression suite — across 16 dedicated parsers covering Articulate Rise, Storyline, iSpring, Adobe Captivate, EdApp, Litmos, Camtasia, PandaSuite, and 10+ other authoring tools. This report establishes the baseline state of the SCORM ecosystem as we enter the second half of 2026.
Executive summary
1. SCORM is alive. Despite predictions of obsolescence, SCORM 1.2 remains the most-uploaded version in 2026, accounting for ~71% of packages we process. Articulate Rise 360 alone represents nearly 29% of all SCORM in our corpus — by far the dominant authoring tool.
2. cmi5 / xAPI growth is real but slow. Only 6% of packages we processed in 2026 use cmi5 or xAPI Tin Can, despite eight years on the market. SCORM 2004 4th Edition is now the second-most-common format, beating pure xAPI by 3×.
3. The "remote launcher" problem is industry-wide.About 12% of packages we process are not SCORM courses at all — they're iframe loaders that point to a remote cloud server (Coassemble, Genially, MathAlea, Elucidat, OpenSesame, LLAB, Scormify, ScormXD). The content lives on the vendor's server, not in the ZIP. This is a growing pattern that breaks SCORM's original portability promise.
4. Authoring tool diversity is high. Our 107-package corpus touches at least 30 distinct authoring tools, with 13 of them represented by a single package. SCORM is a long-tail market with no clear winner outside the top 3 (Rise + Storyline + iSpring = ~50% of the corpus).
Authoring tool market share (2026)
Distribution of authoring tools across 107 real-world SCORM packages uploaded between February and June 2026.
SCORM version distribution
Which SCORM specification do organizations actually use in 2026?
Key trends 2026
Trend 1: Webpack-bundled courses are breaking traditional extractors
Both Articulate Rise 360 and Adobe Captivate have migrated their HTML5 export format to webpack-bundled JavaScript chunks. The content is serialized either as inline Base64 payloads or distributed across chunks. Naive extractors that parse static HTML get nothing back. Modern SCORM converters need dedicated parsers per tool — there is no longer a one-size-fits-all approach.
Trend 2: SCORM is increasingly multi-lingual
In our corpus, 18% of packages are non-English. Russian, Italian, Spanish, Portuguese, and Polish are the top non-English languages. This has implications for downstream tools: word counters that rely on ASCII-only regex word boundaries (the common JavaScript \b bug) silently undercount non-Latin scripts by an order of magnitude.
Trend 3: AI-driven extraction is the new floor, not the ceiling
About 7% of packages we process have content that is impossible to extract without OCR (e.g. iSpring Page slides rendered as PNG screenshots). Gemini Vision OCR at ~$0.002 per slide is now economically viable as a fallback layer. The next twelve months will see OCR + AI transcription transition from optional add-ons to core capabilities for any serious SCORM extractor.
Trend 4: The "cloud launcher" anti-pattern is escalating
SCORM was designed for portable, self-contained packages. Yet 12% of what ships as SCORM in 2026 is just an iframe pointing to a vendor's cloud. When the vendor goes out of business or changes pricing, the "SCORM compliance" vanishes. Organizations should treat this as content lock-in — and prefer authoring tools that produce truly portable SCORM.
Trend 5: SCORM is becoming AI training data
Increasingly, L&D teams export SCORM not for human readers but for RAG pipelines: feeding course content into vector stores so that internal Slack chatbots can answer policy questions, compliance scenarios, and product training questions. Markdown export with GFM tables is becoming the preferred output for AI engineers.
Methodology
The numbers in this report come from a curated "regression corpus" of 107 SCORM packages uploaded to scormconverter.com between February and June 2026. Each package was processed through all 16 dedicated parsers plus the generic fallback. We report:
- Authoring tool detection via 30+ fingerprint rules (path-based, manifest-based, and content-based)
- SCORM versionfrom the manifest's
schemaversionelement - Content ratefrom per-file word-count classification (rich ≥1k, medium 200-1k, partial 50-200, thin <50)
- Language from manifest
xml:lang+ content heuristics
The corpus is biased toward early-adopter users who actively seek SCORM extraction. It overrepresents Articulate Rise (the most converter-friendly authoring tool) and underrepresents proprietary LMS-internal SCORM that never gets exported.
Predictions for 2027
- Rise market share will exceed 35%. Articulate continues to dominate the responsive-course authoring space.
- cmi5 adoption will reach 12-15%.The growing "learning experience platform" (LXP) trend pushes vendors toward xAPI-native content.
- AI-extracted SCORM will outnumber human-extracted SCORM. The bulk of conversion volume will shift from human users to automated AI pipelines feeding LLM training and RAG.
- The "cloud launcher" backlash starts. Procurement teams will start auditing SCORM packages for portability before accepting them.
- Native HTML5 export displaces SCORM 2004 4th Edition. By end-2027, expect SCORM 1.2 + native HTML5 to be the dominant pair.
Want to extract your own SCORM corpus?
SCORM Converter ships 16 dedicated parsers for the major authoring tools. Free during beta — no credit card required.
This report is published annually. Want to be notified when the 2027 edition is out?
Cite this report: SCORM Converter Team. (2026). The State of SCORM 2026 — Annual Industry Report. scormconverter.com/state-of-scorm-2026