The recent wave of AI-led layoffs followed by rapid rehiring tells us something important, but perhaps not what many people first assume.

It does not mean AI has failed. It does not mean organisations should slow down innovation or avoid automation. It means some decisions were made too quickly, with too little understanding of how work actually happens inside a business.

In the main article, AI Did Not Fail — Poor Workforce Strategy Did, I explored why many organisations are now learning that AI transformation is not the same thing as workforce replacement. That distinction matters.

A Careerminds study of HR professionals found that many companies that made AI-led cuts later rehired roles they had removed. More than half said they rehired for previously eliminated roles within six months, and many reported that AI required more human oversight than expected.

To me, this is not a technology story first. It is a leadership story.

When a new technology arrives with speed, scale, and economic pressure behind it, leaders face a difficult balancing act. They must be open to change, but not reckless. They must pursue efficiency, but not hollow out capability. They must respond to market pressure, but not confuse urgency with wisdom.

That is where many AI workforce decisions appear to have gone wrong.

The temptation is understandable. AI promises faster output, lower operating costs, improved consistency, and automation of repetitive tasks. In some areas, it delivers exactly that. But work is rarely just a collection of tasks. Work also contains judgement, context, relationships, memory, interpretation, and accountability.

A person in customer service is not only answering questions. They are reading emotion, managing trust, knowing when to escalate, and protecting the organisation’s reputation.

A software engineer is not only writing code. They are understanding architecture, trade-offs, legacy constraints, product intent, and risk.

An HR professional is not only processing employee requests. They are navigating human complexity, policy interpretation, workplace culture, and sensitive judgement.

When organisations remove people before understanding these layers, they can lose more than capacity. They can lose institutional intelligence.

This is why I believe the way forward is human-in-the-loop AI.

That phrase is often used in technical discussions, but it is also a leadership principle. It means people remain responsible for judgement, oversight, ethics, escalation, and final accountability. AI can assist, accelerate, summarise, analyse, draft, monitor, and recommend. But humans must still decide what matters, what is acceptable, and what should happen next.

This is also why AI literacy is now essential at every level of an organisation.

Executives need AI literacy so they do not make strategic decisions based on hype or fear.

Managers need AI literacy so they can redesign work intelligently, rather than simply cut roles.

Employees need AI literacy so they can improve their own productivity and remain relevant in a changing workplace.

Technical teams need AI literacy at a deeper level so they can build systems that are reliable, monitored, secure, and aligned with business reality.

I am optimistic about AI, but my optimism is not blind. AI will change work. Some roles will shrink. Some will disappear. New ones will emerge. The balance will continue to move. But the organisations that handle this transition well will not be the ones that rush into dramatic cuts and then scramble to rebuild later.

They will be the ones that invest in people before replacing them.

They will ask better questions:

Where can AI genuinely improve the work?

Where do we still need human judgement?

What knowledge would we lose if this role disappeared?

Can this person be reskilled or redeployed?

What governance do we need before scaling this system?

What happens when the AI is wrong?

These questions are not obstacles to innovation. They are what make innovation sustainable.

The real leadership challenge is not choosing between people and AI. It is building organisations where people can work better with AI, where technology improves capability, and where workforce decisions are made with discipline rather than panic.

AI is not the enemy. Poor preparation is.

The best response is not fear. It is literacy, thoughtful implementation, and a commitment to upgrading human capability alongside technological capability.

That is the path I believe more organisations need to take.