Weekly · Cloud & AI security
Pyralink™ Prove It
Practical, sourced notes on cloud and AI security for compliance, IT and security teams, and founders. One email every Monday, written by Michael Adedeji, plus a free Secure AI Toolkit you can use today.
Secure sign-up · Weekly
Subscribe by email
We'll email you each new edition. Unsubscribe any time.
The form above is provided by Beehiiv, our email provider, and your details go straight to them. Read how we use them in our privacy policy.
Every Monday · By Michael Adedeji
Who it is for
Written for the people who have to make AI and cloud use safe, and be able to show it. Each group gets a place to start.
Know what is changing
UK and EU dates, open calls for evidence and guidance in plain English, and the questions to put to suppliers.
Start with: regulation tracker, vendor questionnaire
Fix what matters first
Concrete steps on cloud basics, AI keys, agents and prompt injection, with the primary source for each.
Start with: cloud basics, prompt injection guide
Run it as a programme
A checklist to work through, a playbook for when something goes wrong, and a glossary the whole team can share.
Start with: adoption checklist, incident playbook
Get the basics in place
A short policy you can adopt this week, plain-English advice, and a way to protect yourself and your family.
Start with: acceptable-use policy, personal protection guide
What is in every edition
Six parts, in the same order every Monday. Short enough to read over a coffee, with a link for every claim.
This week in AI and cloud security
The few developments that matter, each linked to its primary source.
One practical fix
One change you can make this week, with the steps to do it.
Regulation watch
UK and EU dates, open calls for evidence and guidance: what changed and what is next.
Tool or template of the week
A free template, checklist or tool you can pick up and use.
Myth vs fact
One common belief checked against the evidence.
Quick checklist
Five or six checks to finish before the next edition.
The free Secure AI Toolkit
Nine free resources for anyone securing AI adoption, from a first policy to an incident playbook. Every page links to its primary sources, prints cleanly, and has a Markdown download. No sign-up is needed to read or use them.
AI acceptable-use policy: a starter template
A plain-English, editable starter template for an organisation's AI acceptable-use policy, with notes on what to change. Free to copy. Not legal advice.
Secure AI adoption checklist
A printable checklist for adopting AI safely: shadow AI discovery, data classification, prompt and data-leak controls, vendor due diligence, access control, logging, human oversight and incident response.
AI vendor security questionnaire
A free questionnaire for assessing the security and data handling of an AI supplier, with what a good answer looks like and the red flags to watch for. Printable and downloadable.
Prompt injection and data leaks: a defence guide for staff
A plain-English guide for staff and IT on prompt injection and data leaks when using AI tools and agents: what they are, what to never paste, what to check, and what builders should do. Printable.
Cloud security basics checklist
A free, printable checklist of the cloud security basics that matter most, for IT teams and small organisations using AWS, Azure, Google Cloud or SaaS. Based on NCSC guidance.
AI incident response mini-playbook
A short, printable playbook for responding to AI-related security incidents, such as data pasted into a public tool, a prompt-injected agent, a compromised AI package or a deepfake payment request. For small and mid-size teams.
UK and EU AI regulation and guidance tracker
A dated tracker of UK and EU AI regulation, standards and official guidance that matters to security, IT and compliance teams. Every date links to its primary source. Last reviewed 31 October 2026.
AI and cloud security glossary
Plain-English definitions of AI security, cloud security and AI regulation terms used by compliance, IT and security teams, with links to primary sources.
Protect yourself from AI scams: deepfakes, voice clones and AI phishing
A practical guide for individuals and families on AI voice-clone and deepfake scams, AI-written phishing, and simple verification habits that work. Printable. Based on UK official advice.
How we keep it honest
Security writing is full of hype. These are the rules we hold ourselves to.
Primary sources first
Claims link to the regulator, standards body or the organisation that published them. Vendor research is labelled as vendor research.
Checked before it is included
Dates and figures are checked against the source. If we cannot confirm something, we leave it out.
Templates are starting points
Every template and checklist says clearly that it is general information, not legal advice, and should be adapted to your organisation.
Corrections welcome
If you spot a mistake, email mike.ade@pyralink.co.uk and we will fix it and say so.
AI keys are cloud keys now.
Three incidents in one week went after the same thing: AI provider keys. Here is what a UK organisation can check before Friday.
Read Pyralink™ Prove It on LinkedIn ↗Questions people ask
Short answers about the email and the toolkit.
Is Pyralink™ Prove It free?
Yes. The weekly email and every page in the Secure AI Toolkit are free. You do not need to sign up to read or download the toolkit.
How often does it come out, and who writes it?
The email is written by Michael Adedeji and is sent on Mondays. You can unsubscribe at any time with the link in any edition.
Is the toolkit legal advice?
No. The templates, checklists and guides are general information and starting points. Adapt them to your organisation and have the important parts reviewed by someone qualified.
Can I use the templates in my own organisation?
Yes. You may copy, adapt and print them for your own organisation. Please do not republish them as your own, and keep the credit to Pyralink™ Prove It.
Where do the facts come from?
From primary sources: regulators, standards bodies, government guidance and the organisations that published the incident or advisory. Every toolkit page ends with its sources, and vendor research is labelled as vendor research.
What happens to my email address?
The sign-up form is provided by Beehiiv, our email provider, and your details go straight to them. We use them only to send the newsletter. Our privacy policy explains the details.
I found a mistake. What should I do?
Email mike.ade@pyralink.co.uk. We will correct it and say so.
Who writes it
Michael Adedeji, founder of Pyralink Innovation Ltd. Pyralink™ Prove It is also published weekly on LinkedIn. Prefer to read it there? Subscribe on LinkedIn ↗
Pyralink also makes PyraLink™ Prove and AI Paste Guard. The newsletter and the toolkit are written to be useful whether or not you ever use them.