All policies

Content Review and Currency Policy

Who is responsible for the accuracy of the course, how often it is reviewed, what forces a review out of cycle, and what record we keep of every change.

Last updated 7 August 2026

Purpose and scope

Deep Space Commodities teaches a field that moves. Missions succeed, fail and get cancelled; launch costs fall; states pass space-resources legislation; the international legal position on owning what you extract is actively contested and unresolved. A course on this subject that is not reviewed is a course that quietly becomes wrong.

This policy covers all 73 video lessons across the 12 chapters, their scripts and transcripts, the on-screen graphics, and the assessment bank that sits behind them. It applies to the course delivered by Shift AI Technology Ltd, trading as Deep Space Commodities.

The short answers

For anyone who wants the substance before the detail:

  • How often is the material reviewed? A full review of every chapter every 12 months, a lighter interim check at six months, and an event-driven review whenever something material changes in the field.
  • Is the process formal or informal? Formal. It is scheduled, it has a named person responsible for signing off each chapter, and its outcome is written down.
  • How are changes implemented? Lesson scripts are held under version control, so every edit is a tracked change with an author and a date. Assessment questions are validated automatically before publication and a failure blocks release. Changes to video are made by correcting the script and re-rendering the lesson.
  • What record of changes is kept? The full version-control history of every script, plus a change log of substantive revisions recording the date, what changed, why, and who approved it. The log is kept for six years and is available to The CPD Standards Office on request.

Who is responsible

Dr Tom James is the course author and lead instructor, and holds final responsibility for the accuracy of what the course teaches. He has spent decades in commodity and energy markets and has written books in the field, and the course is built on that published work, brought up to date where the field has moved on since publication.

No chapter is released or re-released without his sign-off. That sign-off is a specific step in the production process rather than a general assurance: each chapter's scripts are reviewed and approved before the lessons are rendered.

Shift AI Technology Ltd is responsible as the provider for making sure the review actually happens on schedule, for keeping the records, and for acting on what a review finds.

Contributing authors and attribution

Several chapters draw on the work of named subject specialists rather than on the lead author alone. Where they do, the chapter is attributed to the specialist, and the specialist's own position is what the chapter teaches.

  • Chapter 2, on the commercial space race, draws on the work of Tim R. Bowler of BBC News.
  • Chapters 3 and 4, on launch law and technology and on robotic versus crewed extraction, are co-authored with Simon Roper.
  • Chapter 8, on space law, draws on the work of Michael J. Listner.
  • Chapter 9, on the international legal framework, draws on the work of Kamil Muzyka.
  • Chapter 10, on interplanetary trade economics, draws on the work of John Hickman.
  • Chapter 11, on astropolitics and international relations, draws on the work of Bleddyn E. Bowen.
  • Chapter 12, on the economics of Mars colonisation, draws on the work of Robert Zubrin. This is contested territory, and the chapter sets out the competing positions in the debate rather than presenting a single settled view.

This matters for review as well as for credit. When something changes in a specialist chapter's territory, the review of that chapter goes to the contributing author's material rather than being handled generically, and a substantive change to a specialist chapter is not made on the lead author's authority alone.

Where the field is genuinely contested — and on the legal status of extracted space resources it is — the course says so and sets out the competing positions. We treat presenting an unresolved question as settled as an error in its own right.

The review cycle

Three things run in parallel: a full annual review, a lighter interim check, and event-driven review.

  1. 1Full review, every 12 months. Every chapter is worked through lesson by lesson against its script. The reviewer checks that the facts are still correct, that figures and dates are current, that named organisations, missions and legal instruments are described as they now stand, that the argument still holds, and that the chapter's exam questions still match what the lessons teach. Each chapter is signed off or listed for change.
  2. 2Interim check, at six months. Deliberately lighter and targeted at the material most likely to have dated rather than at the whole course: mission status, launch costs and price-per-kilogram figures, treaty and regulatory developments, signatory and ratification counts, and anything the lessons describe as current. A small provider that promises two full reviews a year will end up doing one; a full review plus a focused check is a commitment we can keep properly.
  3. 3Event-driven review, whenever a trigger occurs. This runs outside the calendar and is described below.

The date of the last full review and the date of the next one are recorded in the change log.

Why event-driven review matters on this subject

On many courses an annual cycle is enough. Here it is not, because a single event can make a specific claim in a specific lesson wrong overnight, and waiting up to twelve months to correct it would mean teaching something we know to be false.

The kinds of facts that date on this course are not marginal ones. A rover programme that a lesson presents as the working prototype for lunar water extraction can be cancelled, as NASA's VIPER was in 2024, which changes how that lesson has to be taught. The number of states that have signed the Artemis Accords, or ratified the Moon Agreement, moves. National space-resources legislation keeps arriving; the United States, Luxembourg, the United Arab Emirates and Japan each changed the picture when they legislated. Launch cost per kilogram is a live figure, and it is the number the economics of the whole industry turns on. Mission results — a deflection test, a sample return, a landing — either confirm or undermine the arguments a chapter is built on.

So event-driven review is not a fallback for when the annual review is late. It is the mechanism that keeps the course honest between cycles.

Triggers for an out-of-cycle review

Any of the following requires the affected lessons to be reviewed without waiting for the next scheduled cycle.

  • A significant treaty or regulatory development: a new international instrument, a change in ratification or signatory status, a formal position taken at UNCOPUOS, or new national space-resources legislation.
  • A major mission outcome: a landing, a sample return, a deflection or extraction result, a launch failure, or the cancellation of a programme the course describes as active.
  • A material change in launch economics: a shift in cost per kilogram, or a new vehicle reaching or failing to reach operational service, where the course quotes figures that depend on it.
  • A learner, contributing author or reviewer reporting a suspected error.
  • A pattern in assessment data suggesting a question or the lesson behind it is defective, such as a question almost every learner gets wrong.
  • A complaint or appeal that reveals a fault in the teaching material, escalated from the process in our Complaints and Appeals policy.
  • A change in the accreditation or professional context in which the course is offered.

An out-of-cycle review is scoped to the lessons the event touches, not to the whole course. Its outcome is recorded in the change log in the same way as a scheduled review, including where the conclusion is that no change is needed.

How a reported error is handled

Anyone can report a suspected error by writing to info@spaceeconomy.education with the chapter and lesson, the point in the video if they have it, and what they believe the correct position to be.

  1. 1Acknowledgement within 2 working days, from a person.
  2. 2The claim is checked against the lesson script and the source the script drew on. Where the chapter belongs to a contributing author, it is checked against their material.
  3. 3The report is classified as a confirmed factual error, a matter of interpretation on which the course takes a defensible position, or not an error. We tell the person which, and why.
  4. 4Confirmed factual errors are corrected. We commit to correcting the script and any affected assessment question within 10 working days of confirming the error.
  5. 5Where the correction requires the video itself to change, the lesson is re-recorded and the change is scheduled with that chapter's next render. We will not leave a confirmed error uncorrected in the written material while the video catches up, and we will say plainly when a re-record is pending.
  6. 6If the error affected an exam question, the question is corrected and any learner disadvantaged by it has the affected attempt put right, whether or not they were the one who reported it. That is dealt with under our Assessment, Certification and Academic Integrity policy.

We commit to correcting confirmed factual errors. We do not commit to changing material because someone disagrees with a conclusion the course reaches, where that conclusion is one the named author holds and can defend. Where a reader's disagreement is with a genuinely contested question, the right fix is usually to make the contest clearer in the lesson, and we will do that.

How changes are implemented

Every lesson exists first as a written script. The script, not the video, is the master copy of what the course teaches, and every change starts there.

  • Lesson scripts are held under version control, one file per lesson. Every edit is a tracked change carrying an author, a timestamp and a full record of exactly what text changed. Nothing is edited in place without a history.
  • A changed script is reviewed and signed off before it is used. For a substantive change to a specialist chapter, that means the contributing author's material governs; for everything else, the lead author signs off.
  • Where the change affects what is spoken on screen, the lesson is re-rendered from the corrected script and the new video replaces the old one, along with its captions and transcript. The transcript the AI tutor reads is derived from the same script, so a corrected script corrects the tutor too.
  • Assessment questions are validated automatically before publication. The check confirms that each question has exactly one correct answer, that every question carries an explanation, that no prompt is duplicated within or across exams, and that the correct answer is not systematically the longest option. A failure blocks publication rather than raising a warning.
  • Substantive changes are recorded in the change log. Typographical fixes and production tidying are not logged individually; they remain visible in the version history.

What record of changes is kept

Three records, each doing a different job.

  • The version-control history of every lesson script. Complete, permanent, and precise to the sentence. It answers what changed and when, for any lesson, at any point in its life.
  • A change log of substantive revisions, recording the date, the chapter and lesson affected, what changed, what prompted it — scheduled review, event trigger, or a report from a named source — and who approved it. The log is kept for six years.
  • Review and audit documents. A scheduled review produces a written record of what was checked and what was found, chapter by chapter, including where the finding was that nothing needed to change.

These are not hypothetical. In June 2026 all 73 lesson scripts were audited line by line against the source text they were written from. The audit found and fixed internal contradictions and misread figures in eleven lessons, separated deliberate post-publication updates from errors, and listed the added material that needed the author's confirmation chapter by chapter. The audit document, its findings and the corrections made are retained as part of the record.

The change log, the review records and the audit documents are available to The CPD Standards Office on request.

Keeping the assessment bank current

There is little point correcting a lesson and leaving an exam question that tests the old version of the fact. The assessment bank is reviewed on the same cycle as the lessons and in the same pass.

  • When a lesson changes, the questions drawn from it are checked in the same pass, and any question that depends on a fact that has moved is rewritten or withdrawn from the pool.
  • A question is only ever written against taught material. If a review removes a claim from a lesson, any question testing it goes with it.
  • The automated validation is re-run on the whole bank whenever questions change, and publication is blocked if it fails.
  • Assessment performance is reviewed for defects the validation cannot see. A question that almost every learner gets wrong is treated as a suspected defect in the question or the lesson, not as evidence that learners are not reading it.

The rules governing the exams themselves — pool size, sampling, pass mark, attempts and appeals — are in our Assessment, Certification and Academic Integrity policy, and are not restated here.

How the course is produced, stated plainly

The lessons are presented by a synthetic presenter generated with AI video tooling, using one consistent presenter across all 73 lessons. The presenter is not a likeness of Dr Tom James or of any contributing author. We say so here, on the home page and the public syllabus before anyone buys, and at greater length on our Human Touch page. A learner should not have to work out for themselves that the person on screen is not a person.

What that presenter says is written by people. Every script is authored by Dr Tom James or drawn from the work of the named contributing author for that chapter, and reviewed by a human before it is rendered. The subject-matter judgement in this course — what is taught, in what order, which claims are sound, which questions are genuinely open, and whether a chapter is fit to publish — is human throughout, and it is the part we do not delegate.

AI tools are used in production and on the platform, and the AI tutor is part of the learning experience. The tutor is not a source of course content: it works from the human-authored transcript of the lesson being watched, and it is described honestly, including its limitations, in our Learner Support policy. No AI system decides what the course teaches, rules on whether a disputed fact is true, or signs off a chapter.

Review of this policy

This policy is reviewed every 12 months, and sooner if a review cycle proves to be the wrong shape for the material. The date of the last substantive revision is shown at the top of this page.

Contact

To report a suspected error, ask about a source, or request the change log: info@spaceeconomy.education.

Shift AI Technology Ltd, registered in England and Wales, company number 17159725, trading as Deep Space Commodities. Unit 14, The Maltings Industrial Estate, Southminster, Essex, CM0 7EQ.

This policy is published by Shift AI Technology Ltd, trading as Deep Space Commodities, registered in England and Wales, company number 17159725. Unit 14, The Maltings Industrial Estate, Southminster, Essex, CM0 7EQ.

All policies