A dated log of what changed in Data Goblin: new chapters, corrections, and updated sources, so you can see exactly how the guide has evolved.
When a source changes in a way that affects the guide, we date the change and link the versions we can find. For public promises, we keep the original terms and deadline, and record what the evidence shows was delivered.
November 10, 2026 · New explainer
What the AI hiring studies measured
The Labour explainer compares the age groups, hiring measures and employment estimates in three Canadian and U.S. studies. Each comparison links to its Chapter 16 passage and receipt.
The Data Centres explainer compares electricity bills, site connections and wider grid upgrades across five provinces, with links to the Chapter 8 passages and receipts.
The recaps in seven French chapters now reflect the revised text. In both editions, you can open receipts #23 and #33 from the same opening paragraph of Chapter 5.
Chapter 4 previously gave a ten-year e-waste projection as an annual figure. It now gives the full period, 2020–2030, and separates Canada’s documented recycling share from the global figure. Chapters 6 and 8 now distinguish equipment load, whole-site draw, water consumption and withdrawals.
Chapters 9, 10, 13, 17 and 18 now give the conditions for the Québec and Ontario provisions, including the opportunity to submit observations to someone who can review a fully automated decision. The federal AI Register’s launch count includes projects from research through deployment.
The Stanford scores now identify each edition’s criteria and group, and the submission figures count only invited developers who sent their own report. The minerals finding distinguishes Indigenous lands from the combined land category. Chapter 12 separates CRTC certification from the federal film tax credit.
These are corrections to how the guide read its sources. The receipts show the earlier wording and the evidence used for the correction. The film tax credit keeps receipt #51, and four receipts that were missing from the reader are now available in both languages.
Bill C-22 changed between first reading and the Senate
The March 12 bill proposed up to a year of metadata retention by regulation. The House text of June 18 reduced that ceiling to six months. Our earlier guide described the one-year version. Chapter 10 and receipt #70 now distinguish the two and explain the House text’s safeguards and review process.
The Senate sent the bill to committee on October 22. It was still under study when we checked on November 9. Receipt #70 links the dated texts and proceedings and records the October 22 note that the Electronic Frontier Foundation added to its June article.
Search from the mobile header in English or French to find topics, glossary terms and passages. If search stalls, you can retry without retyping your question.
Use “In this chapter” to jump to a section or share a link. Switching languages keeps you on the same passage.
In Chapter 13’s practical test, select the percentage to see how the researchers counted views. Open receipt #32 for the study’s results, or follow the link back to the passage. We also clarified the six checks in French.