Foundations is the first stage of the Conosco.AI approach: the security, data and governance work a business does before it switches on any AI tool, whether that is an off-the-shelf AI assistant or an app built for the business. It is for growing businesses in London and across the UK that want their AI readiness checked properly, so new tools only surface what each person is meant to see.
AI tools see what your people can see, and they find it faster. A file shared with the whole company years ago and since forgotten can turn up in an answer to anyone who asks the right question. Foundations makes sure that what AI finds is what it should find.
It is also good housekeeping whether or not you adopt AI. Tighter access, cleaner permissions and a clear policy reduce risk on their own.
What does an AI readiness assessment check before you switch on AI?
It checks the three things that decide what an AI tool can reach and how it will be used: who has access, what data they can reach, and what rules apply.
In practice that means reviewing how accounts are protected, finding files and sites shared more widely than intended, checking whether sensitive data is labelled, and confirming whether an AI policy exists. Our AI security readiness assessment covers these checks for businesses about to switch on AI tools.
Why does oversharing matter more once AI tools are switched on?
Before AI, an overshared file was mostly a risk in theory: someone would have to know it existed and go looking for it. An AI assistant does the looking for them. It searches everything a person has permission to open and brings the relevant parts into its answer.
So permissions that were good enough become a real exposure. Salary data, board papers or client contracts can surface in a routine question. Fixing permissions first is the simplest way to stop that.
What we do
- Identity and security baseline. Every account protected with multi-factor authentication (MFA), and admin rights kept to the few who need them. Microsoft 365 and devices are hardened, backed up and monitored, so a new tool does not open a new gap.
- Data permissions and oversharing. We find files shared more widely than they should be, fix the permissions and label sensitive data, so AI tools only surface what each person should see.
- AI policy and governance. Which tools are approved, what they may be used for and who decides. Where you need to show it to others, we align the policy with ISO/IEC 42001.
Who delivers the security work?
Foundations is delivered by Conosco’s security team, the same team that reviews every build before it goes live.
If you already have an IT provider, you can use Conosco.AI alongside them. For the services behind this stage, see identity and access, managed security and Microsoft 365 and cloud.
What does AI governance look like for an SME?
For most SMEs, AI governance does not need to be heavy. It needs to answer a few questions clearly: which AI tools staff may use, what data may go into them, who approves new tools, and who is accountable when something goes wrong.
We write that down in plain language your people will actually read. If clients, insurers or partners ask you to show a recognised approach, the policy can be aligned with ISO/IEC 42001, the international standard for AI management systems. Conosco also offers ISO/IEC 42001 consultancy for businesses working towards it.
How do you measure whether you are ready for AI?
We start with a baseline: how many files are overshared, which accounts lack MFA, which controls are missing. At the end we take the same measures again.
The outcome is simple to state: safe to adopt. You see the overshared files found and fixed, and the controls in place, against the baseline we agreed with you at the start.
Where should you start?
If you are about to switch on a new AI tool, or you are not sure who can see what, start here. We usually start with a one-week AI readiness diagnostic. It needs up to a day of your team’s time, has a fixed fee agreed before we start, and ends with a baseline and a costed plan.
Where this fits
Foundations is the first of four stages, though you can start at any of them. Once your identities, data and policy are in order, the next stage is Adoption: choosing the AI tools that suit each role and making sure people use them well. For Conosco’s wider AI services, see conosco.com/services/ai.