Software & IT services

Keep client code and NDA data out of public AI tools, and show your clients you control AI.

Software, SaaS and IT services teams use AI more than almost anyone. Your clients know it, and their security questionnaires now ask how AI is controlled. We put real controls and evidence in place with Microsoft 365.

Book an AI Data Risk Check How AI leak protection works
Business systems protected by layered data and AI checkpoints.
Where data leaks

How company data reaches AI tools in software & it services.

Most of it is people trying to work faster. That is why the answer is control, not a ban.

01

Client source code

Code from client projects pasted into AI assistants on personal accounts, often against the client contract.

02

Secrets and credentials

API keys, connection strings and passwords pasted along with code to debug it.

03

NDA documents

Specifications, architecture documents and client data uploaded for summaries.

04

Client questionnaires

Enterprise clients ask which AI tools touch their data; vague answers lose deals.

05

SOC 2 and ISO 27001

Auditors expect AI tools to be covered by access control, supplier management and data protection.

06

Many tools, many teams

Each team picks its own AI coding and productivity tools with no central list.

What we put in place

Controls built on the Microsoft 365 tools you already use.

We start with an AI Data Risk Check, then roll out protection in stages so work is not interrupted.

  • Approved AI coding tools. An approved list of AI assistants used with work accounts, with personal accounts blocked on managed devices.
  • Detect code and secrets. Purview patterns for credentials, keys and client identifiers in prompts and uploads to AI sites.
  • Client-ready evidence. An AI register, policy and monthly report you can attach to security questionnaires.
  • Agent governance. Owners, identities and permissions for the AI agents your teams build on client data.
FAQ

AI data protection for software & it services: questions

Can developers keep using AI coding assistants?

Yes. We help you approve the tools you trust, require work accounts and stop code and secrets reaching unapproved tools.

Does this help with SOC 2 or ISO 27001?

It gives you documented AI controls and evidence that map to access control, supplier and data protection requirements. Your auditor performs the assessment.

How do we answer client questions about AI?

With a short AI policy, a register of approved tools and a summary of controls and monitoring. We prepare these as part of the program.

Where should we start?

With the AI Data Risk Check, to see which AI tools teams actually use and what data reaches them.