Choosing an AI tool shouldn't stop finance teams from starting. Learn the engine vs car analogy and how to move safely in an AI world.
Waiting for the perfect AI tool can become a very polished form of procrastination. That was one of the clearest messages from our recent webinar on automation and AI in modern finance functions.
For finance leaders, the question often sounds sensible: should we choose Copilot, ChatGPT Enterprise, Claude, or wait for the next generation of platforms to hit the market?
Scott McLiver, Chief AI Officer at PwC New Zealand, offered a simple analogy: refusing to start because you don't know which tool will win is like walking to work because you're not sure who'll win the Formula 1 championship.
Any reasonable car gets you to your destination faster than walking.
Tool selection matters, but not as much as starting
AI tools are evolving quickly. Capabilities change month to month, and enterprise integrations are improving across the major platforms. That can make the first decision feel heavier than it needs to be.
Most finance teams don't need to predict the permanent winner. They need to choose an enterprise-approved tool, set guardrails, train their people, and start actually using the tools.
If the organisation later moves from one tool to another, many workflows are more portable than traditional software. Switching may still create friction, but it's usually better than standing still.
The engine vs car analogy
The same AI model can perform differently depending on where you use it. McLiver compared this to Formula 1: the Large Language Model (LLM) is the engine, and the application around it is the car.
Put simply, that means using Claude, for example, inside one environment may not feel exactly the same as using Claude directly. The engine may be similar, but the surrounding interface, tools, permissions, connectors, workflow design, and user experience can change the result.
For finance teams, this explains a common frustration. A model that performs well in one product may feel weaker in another, even if the underlying intelligence layer is related. The car matters.
Choosing the right tool is an important decision, but choosing a tool to get started with should be your first decision.
Copilot, Claude, and ChatGPT all have a role
In many organisations, Copilot is the natural starting point because Microsoft is already embedded across Excel, SharePoint, Word, Teams, and Outlook. That matters for security, adoption, and internal approval.
At the same time, tools like Claude and ChatGPT Enterprise can provide a more direct experience of frontier AI capability, particularly for teams experimenting with documents, reasoning, analysis, and human-led agents.
The right choice depends on risk settings, technology environment, use-cases, and governance, but the choice shouldn't become a handbrake.
Finance teams need guardrails before scale
Starting doesn't mean letting everyone use AI however they like.
Finance teams should be clear about which tools are approved, what data can be used, what outputs require review, and which tasks are not appropriate.
That is especially important when dealing with leases, trial balances, financial statements, contracts, audit evidence, and IFRS 16 or AASB 16 reporting. Consumer tools shouldn't be used for business data. Enterprise-grade tools and approved SaaS providers are the safer path.
How to choose without getting stuck
A practical selection process should focus on enterprise approval, data protection, system integration, repeatable workflows, training, and governance. If the basics are covered, the next step is controlled adoption.
AI will keep changing. The leading tool today may not lead in every category tomorrow, but finance teams that build fluency now will be better placed to adapt.
The worst option is waiting for certainty in a market that won't stand still. Choose a safe car. Learn how to drive it. Upgrade when you need to.
Talk to LOIS Leasing about how finance teams can use automation and enterprise-grade systems to reduce manual work and strengthen lease accounting control.