In the AI era, stop managing employees as if they are mere workhorses.

In 2026, the management of quite a number of enterprises has encountered a kind of unspeakable frustration.

Companies are purchasing more and more systems, with increasingly fancy interfaces, and employees can generate a well-illustrated proposal report in just a few seconds. However, when settling accounts at the end of the month, projects that were supposed to be delayed are still delayed, cross-departmental meetings still end up in heated arguments, and no one responds to customers' actual demands in a timely manner. What is more ironic is that many employees have learned to use intelligent tools to generate massive volumes of seemingly professional but actually unread materials, which are specially used to cope with internal assessments.

A survey conducted by Moka AI in the fall of 2026 targeting hundreds of enterprises points out that more than 80% of enterprises have never sorted out their internal division of labor from the perspective of intelligent collaboration, and over 70% of managers admit that they cannot quantify the real tangible returns that new technologies bring to the business at all.

Many team leaders suddenly find that the management rules that have worked perfectly for them for decades are now getting stuck.

Professor Chen Chunhua, a renowned management scholar, raises an extremely thought-provoking question in the preface of her new book *Management in the AI Era*: "If all your management experience is wiped out today, what else can you rely on to steer the future?"

This is by no means an alarmist statement. The edifice built by modern management over the past hundred years is encountering unprecedented loosening. When the old routines completely fail, what real problems are team leaders facing? How on earth should they adjust their current management practices?

Reports are generated a thousand times faster, but decisions still take a month to make

The first problem encountered by enterprises after introducing intelligent technologies is the serious disconnect between superficial efficiency and real benefits.

Professor Chen Chunhua analyzes this "AI gain paradox" in the book: Intelligent software can be seen everywhere in many companies, but no changes can be observed in real profits and turnover rates.

Many management teams are keen to publicly claim that financial software has compressed the report generation time from one week in the past to ten minutes. This is indeed progress, but it can hardly be counted as a business gain. If the report is completed in ten minutes, but the bureaucratic chain of multi-layer reporting is still retained behind it, and an actual expense still needs to go through five levels of approval and can only be finally decided at the regular meeting half a month later, then the few days saved are not worth mentioning at all in the face of long decision-making delays.

This practice is like installing an aero-engine on a wooden horse-drawn carriage, whose heavy axle cannot withstand such impact at all. If the process itself is full of obstacles, blindly introducing fast-paced software will only amplify the originally unreasonable approval and internal friction at a faster speed.

The root of the problem lies in the fact that the planning system that has been prevalent for a hundred years is losing its effect.

In the past, enterprises were used to spending months in the fourth quarter of each year on tug-of-war games, where the sales, finance and production departments tested each other, and finally pieced together a compromised annual business budget. This budget system is essentially a pre-blockade of resources. Once the quota is fixed at the beginning of the year, capital and manpower lose the possibility of flow. If new unexpected business opportunities emerge in the middle of the year, they are often forced to be missed due to no budget quota; those outdated old projects that have been proven obsolete continue to consume a large amount of real money just because they have legal budget.

The plans in the industrial era are based on low-frequency quarterly reviews, while today's market demand changes are calculated in seconds. Relying on the plan written half a year ago to cope with the rapidly changing market is in itself a highly risky act of marking the boat to find the sword.

Professor Chen Chunhua puts forward a profound judgment in the book: What the AI era subverts is not management, but the management that treats people as machines. If the old rules that regard employees as execution gears are not changed, even if the enterprise purchases more computing power, it will only go around in a low-level endless loop.

Still dividing work by positions? The company is forcing employees to become "territory-guarding gears"

The second common pain point lies in the traditional position-based division of labor.

Many companies are used to managing people by relying on detailed job descriptions, which stipulate everything from the scope of responsibilities, reporting lines to assessment standards in an airtight manner. This practice was extremely effective in the past stable large-scale production, because everyone only needed to guard their own workstations.

But today, with highly mature technology, for any repetitive affairs that can be written as fixed steps, the processing speed and accuracy of software far exceed that of human beings.

Professor Chen points out: "Static positions are the last will of the industrial era, while dynamic roles are the admission ticket to the AI era."

When most of the basic work of a position is taken over by tools, the remaining affairs are almost all complex problems that require cross-functional collaboration. If managers still require employees according to the past position boundaries, they are artificially building departmental walls. The marketing department pushes the problem to the operation department, the operation department pushes it to the technical department, and everyone says justifiably that this is not within their scope of responsibility.

The book gives an extremely pungent summary: "Positions are for guarding territory, while roles are for winning battles."

To prove the power of breaking position boundaries, Professor Chen Chunhua records in Chapter 4 the real reform of a cross-border e-commerce Company S (pseudonym) with an annual revenue of over 10 billion yuan. This giant with more than 10,000 employees used to have 3,000 designers and 5,000 copywriters, and implemented a strict internal functional division of labor. From discovering overseas fashion trends to finally putting a new product on the shelves, it had to go through a long relay chain including research, scheduling, translation, legal review and other links, taking an average of 14 days. According to internal company statistics, nearly 40% of the prime sales period was unnecessarily wasted in internal communication.

To save itself, Company S completely demolished the original independent functional departments and established an agile hub orchestrated based on an intelligent system. Instead of recruiting executives with a single skill, it set up full-stack project leads. The system crawls real-time fashion trends on overseas social media, automatically generates design sketches and multilingual copy, and the project leads make screening decisions based on their professional aesthetics. After the restructuring, the new product launch cycle was drastically shortened from 14 days to 4 hours, the hit rate of popular products increased by 3 times, and the overall labor cost decreased by 40% at the same time.

The enlightenment of this case to all managers is very straightforward: The future organization is no longer a pyramid with strict hierarchy, but a network that can be reorganized at any time according to tasks.

Enterprises should learn to abandon those rigid job responsibilities and implement a task-driven role system instead. People move along with tasks, and roles are set up according to scenarios. When a task starts, all personnel are quickly in place, and the team disbands immediately after the task is completed to form a new team for the next task. Only by making personnel flow can the team outrun external changes.

Don't use clock-in software to spy on employees, check who is patching loopholes for the system

After the organizational structure is deregulated, the practical problem that managers are most concerned about emerges: How on earth should assessment be carried out? What exactly should daily management focus on?

Many managers feel extremely anxious because they lose direct visibility into the work process, so they use digital means to monitor employees even more aggressively. They count the number of keystrokes employees make every day, monitor screen activity tracks, and check the hourly online record of software usage.

Professor Chen Chunhua directly gives a warning in Chapter 6 of *Management in the AI Era*: If managers still focus on checking the frequency of employees' keystrokes, that is not only stupid, but also encouraging employees to perform "low-level diligence". When supervision goes to extremes, what you get in return will only be system-wide cheating, and the team's most precious self-drive will be completely stifled.

In the intelligent era, the true meaning of control is to provide navigation to avoid getting lost, rather than building high walls to trap people. If team leaders want to get out of confusion, they must complete three resolute action transformations:

First, transform plans from static prediction to dynamic cruise.

Stop forcing the team to submit an unchanging annual target figure. Professor Chen Chunhua suggests that managers establish a traffic light mechanism similar to autonomous driving. Within the range of daily business fluctuations, the system automatically allocates resources according to data, and managers do not need to interfere with trivial approvals; when core business indicators deviate, the system provides alternative disposal plans for managers to make decisions; only when major events such as sudden industry changes occur, will managers step in urgently and take over the command. This allows managers to get rid of the heavy and trivial daily scheduling and focus on controlling major directions.

Second, transform assessment from counting working hours to rewarding experience sharing.

Since basic work has been greatly simplified by automation tools, the logic of paying salaries based on attendance time must be adjusted. Smart managers should assess experience contribution.

In the case of cross-border e-commerce Company T (pseudonym) recorded in the book, the enterprise decisively canceled the individual sales target, and turned to assessing how many practical and efficient scripts verified by actual combat employees input into the public knowledge base, and how many algorithm loopholes employees corrected in their daily use. Employees soon found that turning the experience they explored into team-shared tools could get long-term system revenue sharing. Everyone voluntarily made their skills public, and the transaction conversion rate of the whole team was greatly improved accordingly. The core of assessment should be who makes the whole team smarter, not who stays at the workstation the longest.

Third, retain the final human judgment line of defense at key decision points.

The system calculates cold numbers and probabilities, but the real business world is full of complex emotions and ethics. In Chapter 5 of the whole book, the author records a scenario about financial risk control deduction: after retrieving massive data, the intelligent system calculates that if 15% of the senior employees judged to have poor performance are laid off, the cost can be directly saved by 1.8 million US dollars every year.

But CEO Lin Feng pressed the pause button before signing. He keenly found that the reason why these senior employees performed mediocre in the system data was that they had long been on the front line maintaining the company's most difficult old customers with the most complex credit situations. Algorithms can easily capture the fluctuations of capital flows, but cannot understand the long-term trust established in face-to-face communication.

Lin Feng did not blindly follow the system's layoff suggestion. Instead, he transformed these experienced veterans into business mentors, teaching the system to identify customers' real psychology in business negotiations, and adding humanistic judgment to the risk control rules. Half a year later, the business teams led by these senior managers saw a 40% increase in the retention conversion rate of high-difficulty customers.

This scenario sounds the alarm for managers: Machines are responsible for efficiency, and humans are responsible for meaning. The system can only find local shortcuts from historical data, and it cannot bear the weight of long-term credit for the enterprise. Managers must act as the moral guide of the team, and add a layer of humanistic care on top of cold computing power data.

Professor Chen Chunhua writes in the book: "Traditional managers control people through processes, while effective managers empower people through roles."

The essence of management never changes, which is always to stimulate potential. When mechanical and repetitive labor is gradually taken over by technology, team leaders can finally get rid of the tedious prison of supervision. Stop treating people as parts, dare to use flexible roles to carry each person's strengths, and maintain the warmth of humanity in the digital age. This is the real confidence for leading a team to go through cycles and win continuously.