AI Could Reshape Middle Management as Goldman Sachs Warns of a New Jobs Era
Artificial intelligence is no longer being discussed only as a technology that could replace repetitive tasks. It is increasingly changing how companies are organised, how employees build careers and who performs management work.
A fresh warning from Goldman Sachs has put one part of that transformation in the spotlight: middle management.
Kevin Sneader, who leads Goldman Sachs' Asia-Pacific business outside Japan, said new finance employees could begin their careers managing AI agents rather than spending years performing the routine analytical and preparatory work that traditionally helped them climb the corporate hierarchy. His comments were made at the Milken Institute Asia Summit in Singapore on October 8.
The implication is significant. If AI systems increasingly handle the work traditionally assigned to junior employees, companies may need fewer layers of managers to coordinate that work. But this is not the same as saying that all middle managers will disappear overnight. The more immediate possibility is a restructuring of responsibilities, with some managers redeployed into new roles involving AI supervision, decision-making and oversight.
What Goldman Sachs Is Actually Warning About
The traditional professional-services career ladder has generally worked like a pyramid.
A new employee starts with relatively routine tasks such as collecting information, preparing presentations, analysing data and building financial models. With experience, that employee becomes responsible for reviewing other people's work, managing teams and eventually making higher-level decisions.
AI could disrupt this progression.
Sneader said young employees at Goldman are already entering a world where they are expected to manage AI agents. Instead of supervising a large group of junior workers, a young professional could potentially coordinate a number of software agents that perform different tasks.
That changes the role of the employee. The important skill is no longer simply completing the analysis. It becomes knowing what to ask an AI system to do, checking whether the output is reliable and deciding what should happen next.
Sneader also said that moving existing middle managers into new roles could become a major generational challenge.
That is a more nuanced warning than the headline "AI will erase middle management" suggests.
Why Middle Managers Are Particularly Exposed
Middle management exists partly because organisations need people to coordinate work between senior decision-makers and large numbers of employees.
AI agents could reduce the amount of coordination required.
Imagine a financial-services team that previously had 20 analysts reporting through several layers of associates and managers. If AI can perform a large part of the initial research, data preparation and document production, the organisation may eventually need fewer people supervising those processes.
Some management responsibilities could move directly to frontline employees who control AI systems.
As Sneader put it, the management task could move away from the middle manager and toward the frontline.
This is why the potential disruption is structural. AI is not simply replacing one job with software. It could change the number of layers a company needs between employees and senior leadership.
The Bigger Problem: Fewer Entry-Level Jobs
The middle-management debate cannot be separated from what happens at the bottom of the corporate ladder.
If AI performs much of the routine work traditionally given to junior employees, companies may reduce entry-level hiring.
That could create a difficult problem for professional services. Junior employees traditionally learn by doing basic work before progressing to more complicated responsibilities. If AI removes much of that early-stage work, companies could find themselves with fewer junior employees today but also fewer experienced professionals several years later.
Other participants at the Singapore discussion raised the same concern. Sandra Peterson of Clayton, Dubilier & Rice said the traditional professional-services "pyramid" may no longer survive in its existing form because AI could reduce the amount of grunt work and therefore the number of entry-level positions required.
This creates what could become the most important employment question of the AI era: if AI removes the first rung of the career ladder, how do workers acquire the experience needed to reach the top?
Goldman Sachs Research Is More Nuanced Than the Headline
Goldman Sachs' own economic research does not predict that AI will simply eliminate huge numbers of jobs in the immediate future.
In April 2026, Goldman Sachs Research found that AI was producing a mixed effect on the US labour market. Jobs in occupations where AI substitutes for human labour were under pressure, while employment was increasing in roles where AI complemented human workers. Its research estimated that AI had reduced monthly payroll growth by roughly 16,000 jobs over the preceding year and increased unemployment by about 0.1 percentage point.
Goldman has also estimated that around 300 million jobs globally are exposed to AI automation over a 10-year period. Exposure, however, does not mean that all those jobs will disappear. It means that some portion of the work involved could potentially be automated.
More recently, Goldman Sachs Research found that AI adoption among major developed economies was generally around 15%–20%, while the labour-market effects remained relatively limited at the economy-wide level. The research also identified younger workers as facing stronger hiring headwinds in some AI-exposed areas.
That distinction matters.
AI exposure is not the same as job elimination.
A job may change substantially without disappearing.
Agentic AI Could Make the Shift Faster
The technology behind the latest discussion is increasingly known as agentic AI.
Traditional generative AI generally responds to individual prompts. AI agents can be designed to carry out multi-step tasks, use software tools, retrieve information and continue working toward a defined objective with varying degrees of human supervision.
If these systems become reliable enough for business-critical work, companies could use them as digital workers rather than simply productivity tools.
That could have a larger impact on organisational structures.
A manager who once supervised ten people might instead supervise a smaller number of employees and several AI agents. In another organisation, a junior employee could coordinate multiple AI systems while senior managers focus on strategy, client relationships and decisions requiring human judgment.
Goldman Sachs' July 2026 discussion on the future of employment also highlighted middle managers as a category where advances in agentic AI could become particularly important, while stressing the uncertainty surrounding how quickly these changes will occur.
What Jobs Could Become More Valuable?
AI disruption does not necessarily mean that management skills become irrelevant.
In fact, some human capabilities could become more valuable.
These include:
Judgment: deciding when an AI recommendation should be accepted or rejected.
Accountability: taking responsibility for important business decisions.
Client relationships: handling negotiations, trust and complex human interactions.
AI supervision: coordinating multiple AI systems and checking their outputs.
Domain expertise: understanding whether an AI-generated answer makes sense in a particular industry.
Leadership: motivating people and setting organisational priorities.
Goldman Sachs research has found that occupations involving more human interaction, higher-stakes decisions and less repetitive tasks can be less exposed to automation.
The likely future, therefore, is not necessarily "AI versus humans." It may increasingly be humans who know how to manage AI versus humans who do not.
What It Means for India and the IT-Banking Workforce
The issue is particularly relevant for India because the country has a large workforce in IT services, business-process outsourcing, financial services and other knowledge-intensive industries.
Indian companies have already been investing heavily in generative AI, automation and productivity tools. If AI reduces the amount of routine work required per employee, businesses could eventually operate with flatter organisational structures.
That could pressure traditional entry-level hiring in some white-collar sectors while increasing demand for workers with AI, data, cybersecurity, product and domain expertise.
For India's young workforce, the lesson may be less about avoiding AI-exposed careers and more about developing skills that complement automation.
The transition could also create new opportunities. Goldman Sachs has argued that AI can create additional employment through productivity gains and new economic activity, while AI-related infrastructure such as data centres and power systems creates its own demand for workers.
The Real Risk Is a Broken Career Ladder
The most important part of the Goldman warning is not that middle managers will suddenly become redundant.
It is that the traditional path from junior employee to senior professional may stop working in the same way.
If AI handles the work that once trained young employees, companies will need new ways to develop human judgment and experience. Training, mentorship, apprenticeship and responsibility could become more important precisely because routine work is becoming less valuable.
For businesses, the challenge will be finding the right balance between automation and human development.
For workers, the safest strategy is unlikely to be competing with AI at tasks AI performs best. It will be learning how to use AI while building expertise in areas where judgment, accountability, creativity and human relationships still matter.
AI and Jobs: What Investors Should Watch
For investors, this trend has implications beyond employment.
Companies that can increase output without proportionally increasing headcount could see productivity and operating margins improve. But excessive automation could also weaken employee development, create regulatory risks or damage service quality if human oversight is reduced too far.
The companies best positioned for the AI transition may therefore be those that can combine automation with stronger human productivity rather than simply cutting jobs.
Goldman Sachs' latest comments are best understood as an early warning about organisational change, not a prediction that every middle-management position will vanish. AI is already changing how work gets done, but the speed and scale of the transformation remain uncertain.
The next phase of the AI revolution may therefore be less about replacing entire professions and more about redesigning the corporate hierarchy itself.
Follow the blog for more updates on AI, technology, business and the future of work.
This article is for informational and educational purposes only and should not be considered investment advice.