Traditional AI and generative AI are both called "AI", but for leaders they behave like different technologies. Traditional AI predicts, classifies or optimises using your historical data, typically for a specific, narrow task. Generative AI creates content, such as text, summaries, images or code, from plain-language instructions, and can help with a huge range of everyday work. Each needs a different strategy.
The key differences for leaders
| Traditional AI | Generative AI | |
|---|---|---|
| What it does | Predicts, classifies, optimises | Creates and transforms content |
| Typical examples | Fraud detection, demand forecasting, recommendations | Drafting, summarising, research, analysis, chat assistants |
| Data needs | Large, clean, historical datasets | Instructions plus relevant documents |
| Who uses it | Usually built into systems | Directly by employees |
| Who builds it | Data scientists, specialist partners | Mostly off-the-shelf tools |
| Speed to value | Months to years | Days to weeks for many uses |
| Main risks | Biased predictions, poor data quality | Confident errors, data leakage, misuse |
| Strategy focus | Projects and data | People, skills and workflows |
Strategy for traditional AI
Traditional AI strategy is largely about choosing high-value prediction or optimisation problems and investing in the data and expertise to solve them.
- Pick specific, measurable problems where better predictions are valuable.
- Invest in data quality and infrastructure.
- Use specialists, in-house or external.
- Manage model risk: testing, monitoring, fairness checks.
Strategy for generative AI
Generative AI strategy is mainly about people and work.
- Approve tools and set clear usage rules.
- Train people widely to use AI well and check its output.
- Redesign workflows around AI assistance.
- Manage risks such as errors, data leakage and over-reliance.
- Measure adoption and outcomes across many teams.
Why the distinction matters
Many organisations apply a traditional AI mindset to generative AI: waiting for perfect data, setting up long specialist projects, keeping it in a technical team. This slows them down unnecessarily. Others apply a generative AI mindset to traditional AI, expecting quick results from prediction projects that genuinely need data work and expertise.
Do you need both?
Most organisations will benefit first and most broadly from generative AI, because it helps with everyday knowledge work across the business. Traditional AI makes sense for specific, high-value problems where you have the data. Some advanced uses combine both.
The leader's role differs too
Traditional AI leadership is largely about sponsoring projects and holding specialists accountable for results. Generative AI leadership is much closer to change leadership: building skills, setting norms and redesigning work across many teams. For a primer on the terms, see AI vs machine learning vs generative AI, and for where agents fit, generative AI vs agentic AI.
Our Leading AI Innovation & Organisational Transformation course focuses on generative AI in everyday work and the leadership skills needed to roll it out across a team or organisation.
Frequently asked questions
Is generative AI replacing traditional AI?
No. They serve different purposes. Generative AI is broadening AI use across everyday work, while traditional AI remains important for prediction and optimisation.
Which should a company start with?
For most organisations, generative AI on everyday tasks, because it delivers value quickly with off-the-shelf tools and builds AI skills widely.
Do generative AI tools need our data to work?
They work with the instructions and documents you provide. Using them well requires clear rules about what data can be shared.