Blackboard plagiarism checker integration.
Add Noplag inside the Blackboard Learn Gradebook via LTI 1.3 Advantage — auto-roster (NRPS), grade passback (AGS), Ultra and Original course views. Unlike SafeAssign it pairs a transparent, auditable corpus with an AI-writing signal that's coming soon; unlike Turnitin the Apache 2.0 engine lets your IT read the detection logic before approving. Free for educators.
Three steps from LTI Tool Providers to a graded submission.
Admin registers the LTI 1.3 tool once. Instructors add it per course from the Content Market or Tools menu. Students submit normally — the integration is invisible to them.
Register the LTI 1.3 tool
From the Administrator Panel → Integrations → LTI Tool Providers → Register LTI 1.3/Advantage Tool, paste the Client ID from the Anthology Developer Portal. Learn generates the Deployment ID automatically — no key/secret rotation. Enable AGS (grade post) and the membership service in Manage Global Properties. Works for the Ultra Experience and Original Course View.
Add it to the course
In Ultra, add Noplag from the Content Market or as a Teaching tool with LTI connection; in Original, place it from the Tools menu in a content area. Toggle it on per assignment and set the threshold (default: flag headlines > 15%, AI-likelihood > 0.7). No Building Block to install, no student-facing change.
Submit as usual
Students submit through the normal assignment workflow — no new accounts, no separate login. The check runs server-side; the report appears in the Gradebook grading view (Ultra) or the Grade Center attempt (Original), next to Bb Annotate. The score posts back on instructor save.
Six things the integration does.
Integration depth matters for FERPA and institutional IT review. We list the LTI surface explicitly so your administrator doesn't have to guess what we do with Blackboard roster data.
LTI 1.3 Advantage
Public 1EdTech spec — your IT reads exactly what we send and receive. Registered once in the Anthology Developer Portal, deployed from LTI Tool Providers. No legacy Building Block, no proprietary extension; tested against Learn's Ultra and Original course views.
Inline in the Gradebook
The report renders next to the submission in the Ultra Gradebook grading view — the same panel placement instructors know from SafeAssign — and in the Original Grade Center attempt. Per-paragraph classification is clickable to source. No new window, no context switch.
Grade passback (AGS)
LTI Advantage Assignment & Grade Services. The plagiarism % posts back to the Grade Center / Gradebook on instructor save — as a column, a feedback comment, or both, with the AI-writing signal joining once it ships. The instructor can override before saving.
An AI-writing signal SafeAssign doesn't have (coming soon)
SafeAssign has never shipped AI-writing detection — Anthology trialled one in 2023 and judged it not fit for purpose. Noplag's AI-writing signal is on the roadmap and rolling out soon, alongside the similarity score, framed as a signal to read, not a verdict to act on blindly. No second vendor, no second contract.
Auto-roster via NRPS
LTI Advantage Names & Role Provisioning Services. The roster pulls from Blackboard automatically — no manual sync. Role-aware: Instructor / TA / Grader / Student are respected per Learn's role mapping, so observers see read-only and students see nothing.
A corpus you can see into
SafeAssign matches against its own proprietary internet index (Anthology's docs note a 1–3 day lag and gaps). Noplag matches against an open, documented corpus — web, academic, and your institutional submissions — with citation-aware exclusions so quoted-and-cited passages aren't counted as plagiarism.
What the integration looks like to an instructor.
The same Gradebook grading view you already use. Noplag appears in the panel beside the submission — the placement SafeAssign uses today — with per-paragraph classification clickable to source and threshold-flagged intervals highlighted in the text.
Why the modernist novel matters
MARIA GONZÁLEZ · ESSAY 3 · 1,632 WORDS · SUBMITTED MAR 14 11:42 PMIn Anna Karenina, Tolstoy writes that “happy families are all alike; every unhappy family is unhappy in its own way” (Tolstoy 5). This opening line is often cited as the moment the modernist novel first announces its psychological method.
The modernist project, however, did not fully take shape until early-twentieth-century writers self-consciously rejected the omniscient narrator in favor of consciousness rendered moment-by-moment as it flows through perception.
UNCITED PARAPHRASE · Eysteinsson (1990), p.96 · click to view sourceVirginia Woolf put it directly: “life is not a series of gig lamps symmetrically arranged; life is a luminous halo.” That sentence, written in 1925, is doing the work the previous century's narrators had refused.
What's different from the SafeAssign you have and the Turnitin you've been quoted.
What your administrator needs before approving the integration.
Blackboard integrations get reviewed by Information Security and the Office of General Counsel before deployment — typically a 4–8 week process at large institutions. We give your reviewers what they need: a completed security questionnaire and a data-flow diagram showing exactly what crosses the boundary. FERPA-compliant DPA: Noplag as a school official under your existing Blackboard institutional FERPA agreement; the only student PII is the LTI-passed identifier, not name + email. GDPR-compliant DPA for European institutions: data-processor designation, sub-processor list, EU residency endpoint (Frankfurt). We're a relaunching open-core company — formal attestations like SOC 2 are on our roadmap, not yet in hand, and we don't claim otherwise. What we offer instead is stronger for a review: the Apache 2.0 engine means your IT reads the detection logic and data flow line-by-line rather than accepting our claims on trust, which most reviewers find faster than the equivalent closed-source review.
Read the institutional-review packetWhat Blackboard admins and instructors actually ask.
- Does it work in both Ultra and Original course views?
- Yes. The integration is LTI 1.3 Advantage, which Blackboard Learn supports in both the Ultra Experience and the Original Course View. In Ultra you add it from the Content Market or as a teaching tool with an LTI connection; in Original you place it from the Tools menu in a content area. With Original retiring at the end of 2026, building on LTI 1.3 means nothing to re-do when your institution finishes the Ultra move.
- How is this different from SafeAssign, which we already have?
- SafeAssign is similarity-only and matches against Anthology's own proprietary internet index plus submitted papers — its own docs note a 1–3 day lag and coverage gaps, and it flags quoted-and-cited text as matches. It has never shipped AI-writing detection; Anthology trialled one in 2023 and decided it was not fit for purpose. Noplag ships an open and documented corpus and citation-aware exclusions today, with an AI-writing signal coming soon. You can run both in parallel during evaluation.
- Do we have to replace SafeAssign?
- No. Blackboard allows more than one plagiarism tool, so you can keep SafeAssign enabled and add Noplag on specific courses or a single college to compare. Most institutions run a semester side-by-side, weigh report quality, AI-signal usefulness, and cost, then decide whether to consolidate.
- Does the score post back to the Grade Center / Gradebook?
- Yes — via LTI Advantage Assignment & Grade Services (AGS). The plagiarism % posts back to the Grade Center (Original) or Gradebook (Ultra) on instructor save, with the AI-writing signal joining once it ships. Configurable as a column, a feedback comment, or both, and the instructor can override the value before saving.
- How honest will the AI-writing detection be?
- AI-writing detection is on our roadmap and rolling out soon — it isn't live yet. When it ships we'll treat it as a signal, not a verdict — the same caution that led several universities to disable Turnitin's AI detector over false positives. Noplag will report an AI-writing likelihood with the reasoning visible, adjust for non-native English writing, and never auto-penalise. The instructor decides. We would rather under-claim than mislabel a student.
- What does FERPA compliance actually mean here?
- Noplag operates as a school official under your existing Blackboard institutional FERPA agreement. The only student identifier that crosses is the LTI-passed user id, not name + email. We don't hold your roster; we hold anonymous identifiers tied to Blackboard's roster via NRPS. A student account deletion triggers fingerprint deletion within 24 hours.
- Will it work with our SSO (Shibboleth / SAML / Okta)?
- Yes. The LTI 1.3 launch happens after Blackboard authenticates the user through your IdP — Learn acts as the OpenID Provider and sends us a signed launch token. We never see SSO credentials, and there's no separate Noplag login for students or instructors. Your existing SAML / Shibboleth / Okta posture passes through unchanged.
- Can our IT actually audit the detection engine?
- Yes — github.com/NoplagLabs/noplag-engine, Apache 2.0. The shipping pipeline (winnowing fingerprints, alignment, report assembly) is readable line by line, and MinHash LSH, vector reranking, and the AI-writing signal join it as they ship. Your IT verifies what is checked, against what corpus, with what thresholds — which is usually faster than a questionnaire-only review of a closed-source tool, because the “we'd need to see the algorithm” step is already answered.
- Are you SOC 2 certified?
- Not yet — we're relaunching as an open-core company, so SOC 2 is on the roadmap rather than in hand, and we don't claim it. What we offer in its place is stronger for a review: an auditable Apache 2.0 engine your team reads directly, an EU-residency endpoint, a signable DPA, FERPA school-official terms under your Blackboard agreement, and a verifiable right-to-erasure SLA.
- Pricing for an entire institution?
- Educator individual: free on the Pro waiver (annual application, no card). Institutional: Enterprise SLA contracts, typically $0–$15K/year depending on FTE and rollout scope. Public records have shown Turnitin charging comparable institutions anywhere from ~$1.79 to ~$6.50 per student — with AI detection as an upcharge — so the saving is usually substantial, and there's no separate AI line item.
- We already use Turnitin — can we migrate?
- Common scenario. Blackboard allows multiple LTI plagiarism tools per assignment, so the two run in parallel during a transition. Most institutions test Noplag on one college for a semester, compare report quality, AI-signal accuracy, and cost, then consolidate or keep both for different course types. Migration playbook available on request.
Install in Blackboard. Free for educators.
5-minute admin install via LTI 1.3 Advantage. Ultra and Original course views. Inline in the Gradebook. Free for educators on the Pro waiver — annual application, no card. Enterprise SLA contracts for full institutional rollouts.