When it comes to deploying AI across your organisation, there’s no longer a single path. According to Gartner, the age of AI built solely in data science labs is over. Today’s AI solutions come from everywhere — embedded in software, brought in by departments, or crafted in-house. The challenge isn’t just choosing between build or buy — it’s blending all three in a smart, secure, and scalable way.
AI is No Longer Just an IT Concern
The most effective AI strategies now pull from multiple sources. Gartner points out that business units are increasingly selecting their own specialised AI tools — from marketing using content generation platforms to legal teams deploying contract-drafting AI. This decentralised trend is changing how we think about AI governance and integration.
To keep up, IT and AI leaders must act as orchestrators — not gatekeepers — of AI. Their role is to enable a safe, coordinated system where AI from all sources can work in harmony.
3 Core Sources of AI Today
1. Embedded AI in Existing Software
By 2026, Gartner expects 80% of software vendors to have embedded AI into their platforms. These updates will enhance everything from ERPs and CRMs to project management tools. It’s a quiet revolution — your current apps may already be smarter than you think.
What to do: Review your software stack to identify upcoming AI-enabled features and assess how they’ll change workflows and user expectations.
2. Bring-Your-Own-AI (BYOAI)
Departments now shop for their own best-of-breed AI solutions. It’s empowering — but risky. Too many disconnected tools can lead to redundant costs, conflicting outputs, and compliance gaps.
What to do: Build processes to vet and track BYOAI solutions, and create channels for departments to align with central IT standards.
3. Built and Blended AI
In-house development still matters — but with a twist. Instead of building everything from scratch, many teams now blend large foundation models (like those behind GenAI) with custom interfaces and data connections.
What to do: Invest in teams who can customise and adapt foundation models, rather than trying to build everything from the ground up.
Don’t Forget Governance
No matter where your AI comes from, it must be governed. Gartner suggests implementing a Trust, Risk, and Security Management (TRiSM) layer to oversee AI safety and compliance. For small AI portfolios, this might include:
- A central AI committee
- An AI ethics team
- Internal communities of practice
For organisations with large-scale AI adoption, automated governance tools — TRiSM technologies — become essential. These tools act as “guardian agents” to monitor access, ensure compliance, and enforce data safeguards in real time.
Key Takeaway:
The smartest AI strategy is a flexible one. Expect to blend embedded AI, department-sourced tools, and custom-built solutions — while maintaining a strong framework for governance, safety, and collaboration.
Based on insights from the Gartner article “Build, Buy or Blend? Deploying AI in Your Organization” by Mary Mesaglio and Hung LeHong (March 2025). Click here to read more: https://www.gartner.com/en/articles/deploying-ai
