Before You Use
This Tool

Every claim, statistic, role description, and clinical detail must be factually true. The AI may produce plausible-sounding numbers or scenarios that do not match your real experience — these must be replaced with your actual figures and circumstances before submission. Submitting fabricated content is professional misconduct under GMC Good Medical Practice and can result in erasure from the medical register.

AI-generated text has identifiable patterns. Experienced GMC assessors can detect unedited AI output. Rewrite every paragraph in your own phrasing — the language, structure, and emphasis should sound like you. The AI provides scaffolding; the voice must be yours.

If you could not elaborate on a claim under direct face-to-face questioning, it should not be in your statement. The Portfolio Pathway submission is a sworn professional document. You must be prepared to expand on, justify, and defend every figure and reflection you submit.

PortfolioMD's fabrication detection identifies content invented by the AI that was not in your notes. It cannot detect inaccurate information you provide as input. You are solely responsible for ensuring that every case count, timeline, role description, and portfolio reference you enter is accurate, verifiable, and honestly represents your clinical practice. Entering false information to generate a misleading statement constitutes professional misconduct under GMC Good Medical Practice.

Every number, percentage, count, and timeframe must be backed by documentation you actually hold — logbook entries, audit reports, MSF feedback, teaching records, appraisal documents. The AI may suggest specific numbers; replace them with your real figures.

The Portfolio Pathway is fundamentally an exercise in self-reflection. Over-reliance on AI generation risks producing applications that lack the personal insight assessors look for. Use PortfolioMD for structure, language refinement, and gap-spotting — not for the reflection itself, which must come from you.

Do not enter any information that could identify a patient — names, NHS numbers, dates of birth, specific dates, unusual conditions, or combinations of factors that could allow identification. You remain bound by your patient confidentiality duties even when typing into an AI tool. Breaches expose you to GMC, ICO, and civil action.

Do not name colleagues, supervisors, or trainees in your reflective content unless they have given explicit consent. Use anonymous descriptors ("the consultant I worked with," "a CT2 trainee"). Named portfolio referees are different — they have agreed to be referees and that role is expected.

Naming hospitals or trusts is usually acceptable, but combining institutional names with specific incidents, criticisms, or unique patient details can create deidentification risk. Be especially careful with safeguarding, end-of-life, and adverse event reflections.

If you reference research involvement, ensure you respect any embargo, sponsor agreement, or co-author confidentiality. Do not include specific unpublished data, embargoed findings, or other researchers' work without their permission.

Text you enter is transmitted to Anthropic's Claude AI service for processing. Anthropic processes the data in the United States. Anthropic does not train its models on API content, but data may be retained for up to 30 days for safety review. This is the same standard for all developers using the Anthropic API.

PortfolioMD encodes guidance from RCoA 2021 Curriculum and GMC Specialty Specific Guidance, but the AI may misinterpret or extrapolate beyond what is in the source documents. Cross-check all guidance against the official RCoA and GMC publications before submission. The AI is not the final authority.

PortfolioMD does not currently store your generated statements between sessions. Copy and save your work to your own device or document immediately after generation. Closing the browser will lose unsaved content.

Your generated statements are unique to your inputs. Sharing them with other Portfolio Pathway applicants encourages plagiarism, which GMC assessors actively look for. Plagiarism between applicants can end multiple medical careers simultaneously.

PortfolioMD encourages naming specific portfolio items — logbook figures, MSF themes, audit titles, teaching counts. Before submission, verify that every named item actually exists in your portfolio. Triangulating against non-existent or exaggerated evidence is professional misconduct.

Statements often include time-bound claims ("over the last 3 years," "since 2020," "during my SHO years"). The AI may suggest plausible timelines that don't match your actual career chronology. Replace all dates and durations with your real history.

Claims about independence, Lead Anaesthetist role, consultant-level work, or supervisory capacity must reflect your actual position at the time the work was done. Overstating your grade of independence is one of the most common reasons for Portfolio Pathway applications being referred or rejected.

Do not include phrasing that implies NHS, GMC, RCoA, or other official body endorsement of your work unless you actually hold that endorsement. The AI may suggest official-sounding language — verify it before keeping it.

PortfolioMD is one tool among many. Mentors, peer review, College Tutor input, and direct RCoA contact remain essential. Treat this platform as a starting point and review aid, not a replacement for human expert support. The RCoA Portfolio Pathway team can be reached at portfolio@rcoa.ac.uk.