AI supports the assessor
AssessIQ suggests possible evidence-to-criteria matches, gaps, questions, summaries and feedback. It is not an awarding body and does not award qualifications or certify occupational competence. The authorised assessor makes and records the final assessment decision, with the centre’s quality and awarding-body controls.
Do not accept or reject evidence, accuse a learner of misconduct or make a significant decision about a person solely because an AI output recommends it. Human review must involve checking the original evidence and being able to change the recommendation.
What the checker can and cannot establish
AI may misread documents, miss context, hallucinate details or map evidence to the wrong criterion. Video processing may use sampled frames and a transcript; it does not establish that every moment of a video has been inspected. A file that cannot be read or processed is not verified. A completion percentage reflects the recorded mapping or review scope, not a guarantee that the whole qualification has been achieved.
Authenticity and possible AI-generation flags are indicators for investigation. They can produce false positives and false negatives and cannot conclusively establish authorship, plagiarism, identity or real-world competence.
Information sent for analysis
Enabled features send the relevant question, criteria and evidence content to OpenAI through its API. Depending on file type, this may include text, documents, images, sampled video frames and audio/video transcripts. Results may be saved with the evidence and review record. The assistant uses the relevant conversation context for its answer.
Only send necessary information the responsible organisation is authorised to share. Do not enter passwords, payment-card details, unrelated personal information or restricted assessment material. No UK-only processing, zero provider retention or particular accuracy is promised. The Privacy Policy explains responsibilities and providers.
Controls and appropriate use
Feature availability depends on the plan and authorised AI controls. A model may be selected according to the task; a model change does not remove the need for human review. Usage controls and commercial fair use do not guarantee that a response will be correct. Follow your organisation’s instructions and provide a route to assessment by a qualified person if a learner raises an AI-related concern.
Errors, concerns and appeals
Learners should contact their assessor or centre, identify the submission and explain the disputed interpretation. The assessor should review the original file, relevant criterion and learner explanation, record the reasoned outcome, and escalate through the centre’s IQA/appeal procedure where needed.
Report a platform error or privacy concern to jernieiq@hotmail.com. Jernie can investigate the service; it cannot overturn an awarding body’s decision. This route does not replace statutory rights or an awarding body’s appeal process.