I do not think AI will follow the Google and Meta playbook
Something I have been noticing
Over the last few months, I have been seeing more partnerships form between AI companies and financial platforms.
Xero working with Claude. Perplexity connecting with Plaid.
At first, it looks like a product story. Better interfaces. Smarter workflows. More automation.
But the more I sit with it, the less I think this is just about improving software.
Where my mind initially went
My first instinct was to compare this to what we have seen before.
Google and Meta built products that people used for free, captured data at scale, and then monetised that data through advertising.
So the natural question is whether AI companies are heading down a similar path.
Will they sit on top of financial data, learn from it, and eventually monetise it in a similar way?
It is a reasonable question.
But the more I think about it, the less convinced I am that the same playbook applies.
What feels different this time
The data is different.
In the Google and Meta world, the value came from attention. Search queries. Clicks. Behaviour.
In this world, the data is not just about what people look at. It is about how businesses actually operate.
Transactions. Payroll. Compliance. Cash flow.
This is not passive data. It is operational.
And that changes the nature of what can be built on top of it.
Where I think the shift is
The way I see it, AI is not just moving closer to data. It is moving closer to decisions.
When an AI system is embedded into an accounting platform, it is no longer just observing activity. It is starting to shape it.
It can categorise transactions. Flag issues. Suggest actions. Eventually, it may execute them.
That is a different position entirely. It is no longer sitting on the outside, analysing data after the fact. It is sitting inside the system where outcomes are produced.
Why that matters
Once you are in that position, the value is no longer just in the data itself. It is in what you do with it.
More specifically, it is in the decisions that are made on top of that data.
That is where this starts to diverge from the Google and Meta model.
Those platforms monetised attention.
This layer has the potential to influence execution.
And execution carries a different kind of responsibility.
Where this connects back to tax
This is where I start to think about tax because tax is already deeply embedded in these systems.
Calculations sit inside software. Data flows automatically. Outputs are generated with minimal manual intervention.
As AI becomes part of that layer, it does not just make the process faster. It starts to influence how tax outcomes are produced.
How income is classified. How positions are interpreted. How issues are identified.
And at some point, those outputs need to be relied on.
The part that I think is often overlooked
If AI is involved in producing those outcomes, the question is no longer just whether the answer is technically correct.
It becomes:
How was that answer produced? Can it be explained? Can it be defended?
That is not a product question.
It is a governance question.
Where I think this leads
The more I think about these partnerships, the more I feel that the real shift is not about data ownership.
It is about control over decision-making.
The system that stores the data is important. But the system that interprets it and turns it into action is where the real value starts to sit.
And that is where things become more complex.
Because once decisions are being shaped inside systems, accountability does not go away.
If anything, it becomes more important.
A final thought
I do not think AI will follow the Google and Meta playbook.
Not because data is not valuable. But because the role AI is starting to play is different. It is moving closer to where decisions are made.
And in areas like tax, where outcomes need to be consistent, explainable, and defensible, that changes what matters.
Not just the data. But the system that turns that data into decisions. And whether that system can be trusted.