Readiness is about more than data quality
Most readiness frameworks cover strategy, data, people, process and infrastructure. They matter. But they skip the question that causes the most damage in practice: data access. AI assistants are only as safe as the permissions underneath them, and in most companies those permissions have quietly grown for years.
The six pillars we assess
1. Use cases and strategy
Which problems should AI actually solve for you, and which are hype? We identify the few use cases with clear value and low risk so you start where it pays off.
2. Data quality
AI on top of messy, outdated or duplicated data gives confident wrong answers. We check where your key data lives and whether it’s fit to be used.
3. Data access and permissions
The security pillar. We look for overshared files and sites, broad groups, anonymous sharing links and stale guest access: everything an AI assistant could surface to the wrong person.
4. People and skills
Who will use AI, for what, and what do they need to know? A short training plan covers useful prompting and, just as importantly, what never goes into an AI tool.
5. Governance and policy
Is there an AI acceptable use policy, an inventory of AI tools including shadow AI, a review step for new tools, and a named owner? This is the foundation of AI governance.
6. Technology and infrastructure
Do your current platforms support the AI features you want with business-grade controls: SSO, audit logs, data residency and training opt-outs?
Why security has to come first
When an AI assistant is switched on, it can search everything the user can access. That’s the point of it. It also means a file shared with “everyone in the company” five years ago is suddenly one question away. Fixing oversharing before rollout is far easier than explaining afterward why an assistant summarized the board pack for an intern.
Quick self-check: are you ready for AI?
- Do you know which AI tools your staff use today, including personal accounts?
- Is there a written AI acceptable use policy that people have actually read?
- Have file and site permissions been reviewed in the last year?
- Are anonymous or “anyone with the link” sharing links under control?
- Do approved AI tools use business accounts with SSO?
- Is someone accountable for AI decisions?
If you answered “no” or “not sure” to two or more, an assessment will pay for itself in avoided exposure.
What you get
A plain-English report scoring each pillar, the specific access and policy fixes to make before rolling out AI assistants, a short list of high-value use cases, and a prioritized roadmap. If you want help with the fixes, the same team can run them as part of an ongoing AI governance program.