Stonetusker Insights · Five-minute checklist
Artificial Intelligence, Machine Learning and Data: 5-Minute Delivery Risk Checklist and Dates to Watch
Eight yes or no questions for teams that train, ship or run models and data platforms. Each one links to the page that explains it, and the dates below show which gaps carry a deadline.
For: Leaders of teams that build model or data platforms, or ship products with models inside them.
Technical and engineering guidance. Please confirm legal obligations with your own counsel.
The short version
- Eight yes or no questions about delivery risk for artificial intelligence, machine learning and data teams.
- Ask your team to answer them in five minutes, from what they know today, with no checks run first.
- A "no" marks a gap worth a closer look. The dates below show which gaps carry a deadline.
Eight questions for your team
- Could someone say which model version serves production, from a stored record instead of memory? Why it matters
- Can the team retrieve, today, the exact training data snapshot behind that model? Why it matters
- Are the evaluation results and the approval behind each model promotion recorded? Why it matters
- Do inference logs record the model version that answered each request? Why it matters
- Do you have a list of the artificial intelligence systems you build or deploy, with an owner and a risk classification for each? Why it matters
- Have all services, functions and pipelines moved off Python 3.10? Why it matters
- Have you confirmed that no workload runs a frozen Bitnami legacy image? Why it matters
- Do all your clusters run a Kubernetes version that still receives patch releases? Why it matters
Reading your answers
- 7 or 8 yes: in good shape. Ask the questions again each quarter.
- 4 to 6 yes: pick the first "no" that has a date attached, and give it an owner.
- 3 or fewer yes: start with the dated items below, in date order.
Dates for your team
| Date | What changes | Details |
|---|---|---|
| From 28 August 2025 | Bitnami catalog change: the legacy repository receives no further updates, and free images are available on the latest tag only.1 | Bitnami images: find and replace |
| 1 October 2026 (passed) | Python 3.10.22 is the final release, and the 3.10 series receives no further security updates.2 | Radar: Python 3.10 |
| 27 October 2026 (passed) | Kubernetes 1.34 is out of maintenance. The maintained releases are 1.35, 1.36 and 1.37.3 | Radar: Kubernetes 1.34 |
| 2 December 2026 | Providers of generative systems placed on the market before 2 August 2026 have until 2 December 2026 to mark generated content.4 | Radar: Artificial Intelligence Act |
| 2 December 2027 and 2 August 2028 | Regulation (EU) 2026/1744, in force since 27 July 2026, moves obligations for Annex III high-risk systems to 2 December 2027 and for Annex I systems (artificial intelligence in regulated products) to 2 August 2028.5,6 | Model lineage and the Act dates |
A note to forward to your team
Machine learning and data delivery risk checklist
Please answer the eight yes or no questions in this note from what you know today. It takes about five minutes.
Reply by [date] with:
1. The number of "yes" answers (out of 8)
2. The numbers of the questions answered "no"
Owner: [name]
Note: https://stonetusker.com/insights/checklist-ai-ml-data.htmlWant a second opinion?
Answer the eight questions with your team first. If you would like a second opinion, reply with two things only: the number of "yes" answers and the numbers of the questions answered "no". Nothing else is needed. Within five working days you receive a short written note on what to do first. It is free and confidential, and a call is optional.
The button opens an email to subeesh@stonetusker.com with the subject and the numbers to fill in. You can also use the contact page, or reply to the email or LinkedIn message that brought you here.
Next page for your situation
- Model lineage and the Artificial Intelligence Act dates: the 30-minute check behind questions 1 to 4.
- Delivery Deadline Radar: every dated item for your archetype, with a check for each.
Sources
- Bitnami: Upcoming changes to the Bitnami catalog
- Python: Python 3.10.22, final release of the 3.10 series
- Kubernetes: Releases and support dates
- European Commission AI Act Service Desk: Article 111
- European Commission AI Act Service Desk: Article 113
- Official Journal of the European Union: Regulation (EU) 2026/1744 (Digital Omnibus on artificial intelligence)
About the author
Subeesh Sivanandan is the Founder and CEO of Stonetusker Systems. He has worked for more than 27 years in release engineering, DevOps, platform engineering & embedded Linux. His career includes serving as the Principal Engineer & Manager of Release Engineering at Stryker, alongside earlier roles at Nokia, VeriSign, and IP Infusion. LinkedIn profile.
Revision history (recent 3 updates)
Version 1.0, 29 October 2026. Every fact on this page is checked against a minimum of two sources each month.
| Version | Date | Change |
|---|---|---|
| 1.0 | 29 October 2026 | First edition |