The headlines coming out of global banking and fintech right now are hard to ignore. HSBC is reducing its workforce by more than 20,000 people. Block, the company behind Square and Cash App, is cutting close to half its headcount. These are not routine restructures. These are decisions that change lives overnight, and they are being made in the name of AI and automation.
What makes these stories worth sitting with is not the scale. It is the timing. In many cases, the productivity gains being chased have not actually materialised yet. Industry data suggests a significant portion of companies reducing headcount in the name of AI are not seeing the ROI improvement they expected. The technology is being implemented. The people strategy is not. And that gap is turning out to be the most expensive mistake a business can make.
The Speed Is the Problem
There is a pattern emerging in how large organisations are approaching AI right now. The board approves the technology investment. The vendors deliver the tools. And somewhere in between, the decision gets made to reduce headcount first and figure out the human side later.
The logic sounds clean on a spreadsheet. If AI can handle tasks that previously required a certain number of people, you need fewer people. Simple. Except it is not simple, because the assumption baked into that logic is that the AI is already working at the level it was promised to work at. In most cases, it is not there yet. The tools are capable. The implementation takes time, iteration, and people who understand both the technology and the workflows it is supposed to improve.
When you remove those people before the system is ready, you do not get a leaner business. You get a slower one with fewer people to fix the problems that the AI has not solved yet.
Productivity Gains Are Not Automatic
This is the hard truth that does not make it into the vendor presentations. AI does not land in a business and immediately produce output at the level it was modelled to achieve. It requires configuration, training data, workflow redesign, and people who know how to use it well. That last part takes months, not days.
The businesses that are seeing genuine productivity gains from AI adoption right now share a common pattern. They prepared their people before they changed their structure. They identified which tasks and workflows were genuinely suited to automation. They ran pilots with the people who owned those workflows. And they gave their teams time to adapt, experiment, and build confidence with the tools before they changed anything structural.
That is a very different process to announcing a headcount reduction and then deploying AI to cover the gap. One of these approaches produces a stronger business. The other produces a faster mess.
The Human and Process Side Is Where Adoption Actually Fails
70% of AI adoption failures are human and process challenges, not technical ones. That figure is not about resistance or attitude. It is about the reality that AI tools work within existing business processes, and if those processes are unclear, inconsistent, or poorly documented before the AI arrives, the AI will not fix them. It will amplify them.
The same applies to the people operating those processes. If your team does not understand what the AI is doing, why it is doing it, or how their role fits alongside it, they cannot course correct when something goes wrong. And things will go wrong, especially early on. That is not a failure of the technology. It is a completely predictable outcome of deploying tools into an environment that was not prepared to receive them.
This is why the job cut first, AI second approach is so costly. You are removing the people who hold the institutional knowledge your AI implementation depends on, before the AI is ready to replace what they knew.
What Business Leaders Should Be Asking Instead
If you are a business owner or leader watching these headlines from HSBC and Block, the instinct might be to ask whether your own organisation should be moving faster on headcount decisions. That is the wrong question.
The right question is whether your people are being prepared. Not just whether they have access to the tools, but whether they understand what is changing and why, whether they have been given the time and support to develop new skills, whether the workflows AI is meant to improve have actually been mapped and reviewed, and whether the people closest to those workflows have been involved in designing what comes next.
A business that answers yes to those questions is in a very different position to one that bought the software, reduced the headcount, and hoped the gap would close itself.
The fear that AI creates for people in the workforce is real and it is reasonable. Behind every announcement about tens of thousands of jobs being cut, there are tens of thousands of people whose sense of security, identity, and financial stability has just been shaken. Business leaders who treat that as a footnote to an efficiency story are making a mistake that goes beyond strategy. They are missing the human cost entirely.
The Businesses That Get This Right Start With People
The companies that will be ahead in five years will not necessarily be the ones that moved fastest on AI. They will be the ones that brought their people with them. That is not a soft sentiment. It is a competitive observation grounded in how technology transitions actually play out.
When a team understands why things are changing, has been given real support to adapt, and knows how to work alongside AI rather than be replaced by it, you have a business that compounds. The AI makes the people more capable. The people make the AI more effective. That is the outcome worth building toward.
The HSBC and Block announcements are a warning, not a model. The scale of those decisions and the speed at which they were made, ahead of proven productivity gains, is exactly the pattern that leads to organisations discovering twelve months later that they cut the wrong people, broke the wrong processes, and have a technology investment that is not delivering what the board was promised.
If you are looking at AI adoption right now, the technology is the easy part. Your team is the hard part, and that is where the real work starts. Not with a headcount decision, but with a conversation about where your people fit in what comes next, and how you are going to prepare them to get there.
Get in touch at www.xsiv.au/#form