In active build · Advisory Board now being established
ReachIntegrity is a patent-pending architecture for verifying student authorship of academic submissions before they are handed in. We are building the first institutional deployments now, guided by an Advisory Board of leaders in the field of academic integrity.
The crisis
Universities are being asked to protect academic standards in an environment where finished text is no longer evidence of authorship. In a peer-reviewed blind-injection study at the University of Reading, AI-generated submissions went undetected by examiners in 94% of cases, and on average earned higher marks than work submitted by real students.1 A 2025 mixed-methods study of university faculty found AI-specific plagiarism policies were rated as meaningfully less effective than traditional plagiarism policies (28% versus 49%).2
AI detectors offered a tempting answer, but they have not proven reliable. Several leading institutions, including Vanderbilt (an early adopter in 2023),3 have publicly disabled AI detection in their integrity workflows, citing reliability concerns and disproportionate false-positive rates against non-native English writers.4 Where institutions have acted on detector outputs, courts have begun to push back. In January 2026 a New York court annulled an AI-detection-based plagiarism finding against an Adelphi University student, ruling the determination “without valid basis and devoid of reason” and ordering the university to expunge it from his record.5
Applying unreliable AI detection leads to a broken workflow. Students who did the work feel under suspicion. Faculty become investigators. Integrity offices inherit cases that are difficult to prove and painful to resolve. Institutions need a workflow that protects academic standards without forcing faculty to defend uncertain AI detector outputs as decisive proof. They also need it built with their workflows in mind, not handed to them as a black box.
Most integrity tools, including detectors, plagiarism checkers, and stylometric analyzers, operate after the work has been submitted. The institution receives the work, then has to determine whether a human wrote it, then has to act on that determination. That sequence is the source of many of the disputes integrity offices are now being asked to resolve.
ReachIntegrity changes the sequence. Verification happens before submission. Faculty receive verified work instead of uncertain suspicion.
How the architecture works
ReachIntegrity is built around a closed-loop verification method: document, question, verify, before the work is submitted.
Step One
A Microsoft Word add-in captures a tamper-evident record of how the work comes into being.
During authoring, only cryptographic hashes of the document state are transmitted, never the content itself. Students’ drafts remain private to them. What we hold is a sealed fingerprint of the writing process, secured by a hash chain.
Step Two
When the student is ready to submit, the system analyzes the private writing process record and generates an interview tailored to their specific writing journey.
The questions reference the structural evolution of the document itself, utilizing the private writing process record, which cannot be previewed by the student. The questions thus test lived authorship rather than surface-level writing style. The questions may also test the student’s understanding of the content, aiding the learning process.
Step Three
The student sits a structured oral interview sized to the stakes of the assessment. The interview is conducted by an AI interviewer using text-to-speech and speech-to-text, with accommodation pathways available.
The outcome of each verification fits one of four categories: Verified, Failed, Inconclusive, or Unverifiable. Each category carries a reliability score and a signal-breakdown report. Inconclusive outcomes are routed to human review rather than forcing a binary decision that the evidence cannot support.
ReachIntegrity
Legacy detection workflow
The shift
Most integrity tools begin after the work has already entered the institution. At that point, every concern becomes a case: someone must investigate, accuse, defend, appeal, and decide. Faculty take on forensic judgment. Integrity officers inherit adversarial cases. Students experience the process as accusation. The institution carries the consequences either way.
ReachIntegrity is designed to move the authorship verification upstream. Before submission, students complete structured verification based on their own writing processes. If the work is verified, it enters the submission pipeline with a certificate and audit trail. If it is not verified, the institution handles it according to its review or resubmission policies. The institution is no longer forced to build its position on a detector-based accusation, and the integrity event has not yet become a case.
This is not a claim of legal immunity. Institutions still design their policies, determine consequences, and still make decisions about edge cases, but the evidentiary posture changes meaningfully. Faculty receive work they can trust at the moment they receive it. Integrity offices spend their time on policy and accommodation, not on defending probabilistic accusations.
The goal is not to catch students out. It is to make student work easier to trust, and undisclosed AI substitution harder to submit.
Calibrated to stakes
Detectors apply the same statistical scan to a freshman discussion post and a doctoral dissertation. ReachIntegrity is designed differently: verification rigor and validation thresholds can be set to match the length and consequence of each assessment.
Low stakes
A brief check that stays light and low-friction, appropriate for assessments that warrant integrity assurance without slowing the cadence of weekly coursework.
Summative
A longer, more thorough interview. Probes process and structural decisions, sized to a piece of work that materially shapes a grade.
High stakes
The full rigorous protocol: a sustained interview that probes process, subject understanding, and substantive depth. The depth of the verification matches the weight of the credential.
Integrity architecture
Detection-based tools share a single fragile assumption: that you must be highly confident before you act, because acting on weak evidence creates serious damage to a real student. That assumption is what makes detector false-positive rates litigation-relevant. It is what forces institutions to defend probability scores in court.
ReachIntegrity is designed to err in the right direction.
The first response to a verification that doesn’t quite clear a student’s work is procedural: typically support, rework, resubmission, accommodation, or review. The architecture is designed to err on the side of giving the student another path, not on the side of an early adverse mark.
The integrity of a Verified outcome is meaningful and durable. The issuing authority retains the right to revoke a certificate if fraud is later substantiated. Verification is forgiving when it errs in the moment, and serious about the integrity of what it certifies over time.
Who it serves
Students who do the work should not have to write under suspicion. The architecture is designed to give them a structured way to demonstrate authorship and have their submissions certified. No detector scores. No defensive AI humanizers. No anxiety about a statistical guess.
Faculty did not sign up to be forensic investigators. ReachIntegrity is designed to keep forensic authorship review out of the classroom and give teaching time back to faculty. Verification is conducted automatically, before the work reaches them.
Replace post-submission investigation with pre-submission verification. Each Verified outcome carries an audit trail. Each Inconclusive outcome is routed to human review. Each Failed outcome is documented with the signals that produced it, in language that can be explained to a student, a parent, or counsel.
A degree carries more trust when its high-stakes assessments arrive with verifiable evidence of student authorship under the policy the institution sets. As the ancient proverb puts it, a good name is more desirable than great riches. The reputation of a university is not decoration; it is the very asset in which students are investing.
Architecture, not detection
ReachIntegrity is built around a patent-pending closed-loop verification method: document the writing process privately, generate questions from that process, verify the student’s answers against the same record, and allow only verified work to be submitted. The architecture is protected by a tamper-evident record, layered defenses against adversarial input, and explicit outcome categorization that includes honest reporting of uncertainty.
This is not a repackaged detector. It is a distinct architectural approach to the problem of authorship verification.
Where we are now
The patent is filed. The architecture has been designed. The core verification capability is actively being built: writing-process documentation, question generation, an AI-conducted interview, and a four-category outcome. First live deployments are targeted for early 2027. If building needs more time, we will move the date rather than ship a poor first deployment.
Verification at institutional scale should not be designed in isolation. The Advisory Board puts working academic integrity leaders in the room while decisions are still open. Members see development early, shape what gets built, and secure preferential long-term pricing for their institutions when ReachIntegrity launches commercially.
Built today
Patent filed. Architecture designed. Advisory Board open.
In active build
Core verification capability: writing-process documentation, question generation, AI-conducted interview, four-category outcome. First live deployments targeted for early 2027.
Next to build
Richer signal-breakdown reporting, expanded LMS integration, an institutional dashboard, and additional writing environments, prioritized with the Advisory Board.
Advisory Board
The Advisory Board is a small group of academic integrity officers, deans, faculty, and assessment leaders who advise on the build of ReachIntegrity. It is advisory, not governance, and membership of the board carries no institutional commitment.
The offering
Pre-submission authorship verification for student assessments, built into the institution’s existing workflows.
The core capability is what first deployments deliver: documenting the writing process, generating process-grounded questions, conducting verification interviews, and producing certified outcomes. Capabilities beyond this core arrive as they are built, prioritized with the Advisory Board.
FAQ
A small group of academic integrity officers, deans, faculty, and assessment leaders who advise on the build of ReachIntegrity. Members receive build briefings and roadmap previews under NDA, shape build priorities, and secure preferential long-term pricing for their institutions at commercial launch. It is advisory: no governance role, no institutional commitment, no cost.
Nothing. There is no fee, no purchase obligation, and no procurement process. What’s asked is a 30-minute introductory conversation and periodic feedback as the platform takes shape.
Through the 30-minute conversation. We are looking for people close to the problem: integrity officers, academic leaders, faculty, and assessment administrators whose contexts can inform the build. Seats are confirmed on the basis of mutual fit.
The patent is filed. The architecture is designed. The core verification capability, including writing-process documentation, question generation, the AI-conducted interview, and the four-category outcome, is in active development. First live deployments are targeted for early 2027.
Influence and price. Build priorities are being set now, and board input carries the most weight before those decisions are made. Members also secure preferential long-term pricing for their institutions at commercial launch.
ReachIntegrity currently runs as a Microsoft Word add-in. Students write in Word as normal. No proprietary editor is required, and no content ever leaves the student’s device unencrypted. Support for additional writing environments may be added based on Advisory Board consultation.
ReachIntegrity is designed to support FERPA, GDPR, UK GDPR, and POPIA-aligned deployments. Calibration to your specific regulatory context, including data residency, retention controls, access controls, and accessibility accommodations, is part of each deployment’s design work. Compliance documentation, the Data Processing Agreement, and the Accessibility Statement are produced during that work.
Fully supported by the architecture. The institution sets what is allowed, including grammar checking, brainstorming, outline help, and even disclosed AI drafting, and the system verifies authorship under that policy. The point is not to ban AI; rather, it is to verify student authorship under whatever policy the institution sets.
More detailed questions? Bring them to your introductory conversation. The full Buying Committee FAQ is shared with Advisory Board members under NDA.
From here
AI has changed what submitted work can prove on its own. Institutions now need a different sequence: one that protects students doing the work, supports faculty, and gives academic leaders evidence they can trust at the moment work arrives.
Advisory Board members are shaping the standard by which student authorship will be verified for the next decade. There is a seat for someone from your institution.
Join the Advisory Board