How the AI Bust Could Affect Middle Management and Decision-Making
The term “AI bust” usually refers to a potential collapse or sharp correction in the artificial intelligence boom — for example, if AI investments fail to deliver expected returns, AI company valuations fall sharply, or businesses discover that AI is less profitable than they anticipated.
For middle managers, an AI bust could be more than a technology issue. It could change how companies allocate resources, how managers make decisions, and what leadership expects from them.
The interesting paradox is this: even if the AI boom collapses, AI could still permanently change the way businesses make decisions.
1. How an AI bust could affect middle management
1. Pressure to justify AI spending
During an AI boom, managers may receive budgets for AI tools, automation projects, and experimentation. If the boom reverses, senior executives may demand evidence of measurable returns.
Management consequence: Middle managers must justify investments with cost savings, productivity gains, customer outcomes, or reduced risk — not simply because AI is fashionable.
2. Changes to managerial roles
AI can automate parts of reporting, scheduling, data analysis, and routine coordination. If companies reorganize around these capabilities, some traditional management responsibilities may shrink.
Management consequence: Managers may need to spend more time coaching people, resolving conflicts, coordinating across departments, and making decisions that require judgment.
3. Higher expectations with fewer resources
A company might retain AI-assisted workflows even while cutting technology budgets or reducing staff.
Management consequence: Middle managers could be expected to deliver more with smaller teams, while still being accountable for quality, deadlines, and employee well-being.
When AI recommendations produce poor results, executives may ask who approved the decision and whether the risks were assessed.
Management consequence: Managers need to understand AI limitations, verify important outputs, and document why a recommendation was accepted or rejected.
A market correction could separate useful AI applications from expensive experiments that never deliver sufficient value.
Management consequence: Managers who can distinguish genuine business opportunities from hype may become more valuable to their organizations.
These are plausible scenarios, not guaranteed outcomes. The effect would depend on the severity of the downturn, the company’s financial position, and how deeply AI has already been integrated into its operations.
2. How an AI bust could change the decision-making process
The biggest change may be in how managers decide what information to trust and how much responsibility they retain.
Before an AI bust: The AI-first mindset
Collect business data -> Sales, costs, customers, operations -> AI analyzes and recommends -> Predictions, rankings, suggested actions -> Manager approves and implements -> Potential risk: accepting recommendations without sufficient challenge
After an AI bust: The evidence-first mindset
Define the business problem -> Compare AI insights with evidence -> Check assumptions, costs, data quality and alternatives -> Exercise managerial judgment -> Consider employee knowledge, customer context and risks
Make, explain and review the decision
This is a conceptual comparison, not a prediction that every company will move through these exact stages. AI can improve decisions, but overreliance on automated recommendations can also introduce errors and bias. The US National Institute of Standards and Technology (NIST) emphasizes clear human responsibilities and appropriate oversight in AI-assisted decisions.
Four areas of decision-making that could change
Decision-making area | Potential change after an AI bust |
|---|---|
Strategic decisions | More scrutiny of AI investments, expected returns and long-term commitments |
Operational decisions | Greater emphasis on reliable processes and proven efficiency improvements |
People decisions | More caution about using AI to evaluate employee performance, allocate work or recommend staffing changes |
Risk management | More attention to accuracy, accountability, data privacy and human review |
3. A practical example: A middle manager deciding whether to hire more staff
Imagine you manage a customer service department.
Your team is struggling with a growing number of customer enquiries. An AI tool predicts that automation could handle 60% of the workload, so senior management asks you to reduce the hiring budget.
What happens next?
Scenario A: Blind reliance on AI
You accept the prediction and reduce hiring. Later, you discover that complex customer complaints still require human attention, and service quality suffers.
You examine the types of enquiries, test automation on a limited scale, measure resolution quality and cost, and consult your team. You then recommend an appropriate mix of automation and human staffing.
In Scenario B, AI still helps. The difference is that the manager owns the decision rather than simply forwarding the AI’s recommendation.
If an AI bust occurs, this second approach becomes particularly valuable because organizations may have less money available for experimentation and less tolerance for costly mistakes.
4. Will middle managers become less important or more important?
Both outcomes are possible.
Managers at risk
Primarily pass information up and down the hierarchy.
Produce routine reports without adding much interpretation.
Follow procedures without questioning their suitability.
Add administrative layers without improving decisions.
Managers who can adapt
Translate executive strategy into practical action.
Interpret AI outputs using operational experience.
Resolve conflicts and coordinate teams.
Balance financial targets with customer and employee needs.
Take responsibility for decisions and outcomes.
Recent management research has highlighted a related problem: executives may have high expectations of AI, while middle managers encounter practical difficulties implementing it. AI adoption can also add oversight and coordination work instead of simply eliminating tasks.
An AI bust could intensify these tensions, especially if executives expect the same productivity gains while reducing budgets.
5. A new decision-making framework for middle managers
I would recommend a simple five-step approach that works whether the AI boom continues or reverses.
Define the decision. What problem are you solving, and what does success look like?
Use AI as an adviser. Ask it to analyze data, identify patterns, compare options and challenge your assumptions.
Verify the evidence. Check the source data, assumptions, financial impact and practical constraints.
Apply human judgment. Consider organizational context, employee experience, customer needs and consequences that the data might miss.
Measure the outcome. Compare actual results with expectations, learn from errors and change course when necessary.
The principle is straightforward: AI should improve the quality of a decision, not become a substitute for understanding the decision.
6. The bigger picture: What happens to middle management?
My assessment is that an AI bust could accelerate a transformation that was already underway. It would not necessarily eliminate middle management, but it could expose the difference between managers who mainly coordinate information and those who create value through judgment, communication and execution.
The real challenge is a three-way tension:
Senior executives want productivity, cost savings and returns on investment.
Middle managers must turn those expectations into workable decisions.
Frontline employees experience the actual consequences of those decisions.
Middle managers sit between these groups. They can become the organization’s most important link between technological ambition and operational reality — or become squeezed between unrealistic targets and limited resources.
For further reading, the National Institute of Standards and Technology provides a useful foundation in its AI Risk Management Framework.
One final thought: The strongest middle manager in a post-AI-bust environment may not be the person who uses the most AI. It may be the person who knows when to use it, when to question it, and when the situation requires a human decision.