"The agent keeps working all night": how the work week is changing
A century ago, Henry Ford established the five-day work week. Today, technology and AI are dismantling these structures, turning work into a 24/7 process where AI agents operate under human oversight.
This year marks the centenary of one of the moves that shaped the modern world of work. In 1926, Henry Ford introduced a five-day work week, eight hours a day and 40 hours a week in his factories, and the model became a standard adopted later around the world. Today, this standard is gradually eroding. Technology, COVID-19, hybrid work, a new generation of employees, and now AI are breaking down the boundaries upon which it was built: when people work, where they work, how many hours they work, and even who, or what, does the work.
In materials published globally ahead of the model's centenary, the breakdown of the workday into "micro-shifts" is described — short segments of work scattered throughout the day and adapted to the task, the family, and the hours when the employee is more focused. However, this flexibility has two sides: it allows employees more freedom to choose when and where to work and to adapt work to life, but at the same time it blurs the boundary between them. And now artificial intelligence enters this change and undermines an even more fundamental assumption: not just when and where humans work, but how much of the work a human needs to perform at all. AI tools are already capable of taking on some of the routine tasks, and agents can continue to work when the human employee goes home. If Ford tried a hundred years ago to set a limit on the time a person works, the technology of 2026 allows work not to stop at all.
Sharon Steiner, Chief Human Resources Officer at Fiverr, watched this change almost from the beginning. She started her career at IBM, continued to Galileo and startups, and in 2012 joined Fiverr.
"The change doesn't happen in one day. It sounds like 9:00-17:00 gave way to something very amorphous, following technology and after COVID-19. First of all, there are many places that do maintain these working hours, and beyond that, many young guys even work fewer hours a day, and there is a tendency to create for themselves a number of jobs. People today are looking for multiple occupations, and many employees, not just young ones, work at one workplace and have another project on the side, freelance jobs, or are setting up ventures and projects for themselves. Therefore, they limit their working hours to very defined hours."
According to her, the multiplicity of occupations stems from the lack of security that characterizes the current labor market. "They are looking for employment stability. The world today is very unstable, not just in high-tech. Employees can arrange their work at other hours. It is common for an employee to arrive, work from nine to three, go home to the family, and open a computer at eight in the evening to complete a few more hours. It is also natural, because there is globalization, and when working with other countries, one needs to work in the evening, and this comes at the expense of the afternoon."
If once the aspiration was to create a balance between work and life, Steiner argues that the separation itself no longer describes reality. "In our worlds, there is no longer a work-life balance, everything is mixed. I call it life blend. You need to reach a goal, and the employer cares less when and how it happens. At the same time, we are very strict about physical meetings and do not work completely hybrid. We believe that people are social animals, and ideas and things happen out of meetings — corridor conversations, meetings, and not everything can happen through Zoom."
Sharon Steiner notes:
"What is important today is less the experience or the title of the employee, but more the output. I need an employee to reach a certain output, I care less how it happens."
Since the COVID-19 days, everything has really gone wrong, and now it is starting to balance out, and the 9:00 to 17:00 has become flexible. If the dosage, like ours, is 70%-80% in the office and the rest at home, that's excellent.
Resource management instead of HR
Hybrid work and flexibility in hours are already almost yesterday's news. The big change happening now is the entry of a new type of "employee" into the organization: the machine.
"We are lucky to be inside the change where resources are becoming not only human resources but machine resources. AI is entering, and it is no longer human resources management but resource management."
Steiner reviews the development of the HR profession as a mirror of the changes in the market: from dealing with wages and working conditions a hundred years ago, through employee rights, talent management, and employee experience, to the hybridity and flexibility that COVID-19 brought. Now, according to her, another stage is beginning.
"I am Chief Human Resources Officer (CHRO), and lately I call myself Chief Resource Officer, because the agent has been added, and suddenly the employee has someone who works for him 24/7. The AI does not rest for a moment, and for developers, the human has become the weak link because he needs to eat and sleep and have time off, but the agent continues to work. The developer or analyst can go home, and the agents continue to work all night, and what remains for the human is to manage and coordinate these resources. Human judgment is not redundant — it must be inside, and there are new professions. The change is that the work continues 24/7 under the employee's management."
Attempts to break down the full and rigid job preceded Zoom and AI. She returns to the job pairing model, where two people share one job and joint responsibility for the role: "In the 60s in England, job pairing stemmed from the fact that women on maternity leave, who did not want to lose their workplace, would share a role and salary with another woman so they wouldn't be fired. That is, hiring a freelancer."
Today, according to her, the place of the human partner for the job can to some extent be filled by an agent as well. "If you do it right, and there is the right architecture of resources, you can do wonders, it's magic. We are a bit of a 'talking generation' because we haven't finished moving to the future, and much is still rooted in the past. But the AI natives generation that will enter the market in two years will completely influence work. I predict that in less than a decade, robots will also enter the picture."
At Fiverr, they already tried to take this idea into one of the most human processes in the organization — a job interview using AI, but for now, they gave up until they reach maximum accuracy. "I think that in the future, the candidate will have an agent that he will train to speak in his language, and the whole beginning of the process will be agent against agent. It is clear that in the end, there will be a face-to-face meeting, but it is possible to streamline for the basis of the first questions. The agent learns our style."
The productivity economy
So why wouldn't a company give up almost completely on permanent employees and rely on freelancers?
"It is not right for a company to employ only freelancers. There are legal, financial issues, the need for knowledge, experience, and understanding of product capability. A company wants to have talent that will be the anchor. In most companies, there is the core and there is the project-based part. Especially now with agents, it is important to have very experienced, very 10X (effective or having productivity significantly higher than an average employee) and talented company employees who can learn on their own, move from topic to topic, and be able to work across the board. What is important today is less the experience or the title of the employee, but more the output. I need an employee to reach a certain output, I care less how it happens. If there is an employee who decides to recruit freelancers, to use agents for achieving the goal, I have no problem with that."
Studies are already starting to give a numerical expression to this change. An OECD review from 2025, which examined experimental studies on the impact of generative artificial intelligence at work, found an average improvement of 5% to more than 25% in productivity in areas such as customer service, software development, and consulting. The greatest improvement was recorded precisely among less experienced employees or those with lower skills. In one of the studies, for example, the use of AI increased the productivity of customer service employees by 14% on average, but among new and less skilled employees, the improvement reached 34%."
But while organizations are moving to measure output, adopting flexibility, and using agents that work 24 hours a day, employment contracts and a large part of the regulation are still built around human working time. "I assume it will take time until these things change too. In practice, companies are flexible, they no longer punch a clock."
Attorney Hadas Rakach-Dvir, head of the labor law department at Shibboleth, sees this gap as part of a long history in which the law was required time after time to adapt to changes in the way people work.
"If you look at the last hundred years through labor law, you can identify four central milestones, each of which reflects a deep change in the labor market. The first milestone was the recognition that working time must also have a limit, with the establishment of the eight-hour workday model in the industrial world. The second milestone was the transition from protecting the employee in the narrow sense to also recognizing family life, parenthood, and the need to allow broader and more equal participation in the labor market. The third milestone was the recognition that flexibility is no longer an exception, but part of the way the labor market works. Therefore, labor law was also required to deal with more complex and flexible structures of working time, and in parallel, the process of reducing the scope of working hours continued. The fourth milestone, where we are today, is the transition to a digital, hybrid, and data-based world of work. Here, the change in the market largely precedes legislation, so the law develops mainly through the adaptation of existing principles to new technologies, alongside regulatory guidelines and legislative updates.
For example, Amendment 13 to the Privacy Protection Law, which entered into force in August 2025, strengthened the framework for protecting personal information and imposed, among other things, an obligation to appoint a privacy protection officer in a number of organizations. In this sense, labor law does not develop in a straight line, and in most cases, it is not the regulator who leads the change. Usually, the market changes first, and the law is required to catch up. This is especially prominent today, when technology, and in particular AI tools, advances faster than the pace of regulation. In the European Union, a first step has already been taken, the AI Act, which was enacted at the beginning of August 2026 and constitutes the first comprehensive regulatory framework that regulates the development, implementation, and use of artificial intelligence systems, defines AI systems used, among other things, for recruiting employees, promotion, firing, task distribution, monitoring, and performance evaluation as high-risk systems due to their possible impact on career, livelihood, and employee rights.
The legal questions of the new labor market are no longer limited to how many hours a person works, but also who — or what — makes decisions about him, how much is the employer allowed to follow him, what information is allowed to be collected about him, and how to maintain equality, fairness, and privacy in a work environment that is changing rapidly. Therefore, courts and regulators play a significant role today in adapting old legal principles to a new reality."
Attorney Hadas Rakach-Dvir notes:
"The legal questions of the new labor market are no longer limited to how many hours a person works, but also who or what makes decisions about him."
Tools for the future
Even without waiting for regulation, flexibility did not necessarily make employees more efficient. "The discipline of the person is important," says Steiner. "Once there is excessive flexibility, many people lose concentration, and it hurts. I don't like people working from home, I prefer them to come, work, and go about their business. Parents, for example, 'tick' the day, because when there is a deadline, they do what is needed. This doesn't mean that in the evening they don't open a computer for completion."
The big change is no longer just about the question of when and where the employee performs his work, but who, or what, needs to perform it in the first place. At Fiverr, they decided to turn this question into a work tool. In recent months, Fiverr's HR team has been working on a tool that has become part of the company's recruitment processes. HIP (Hiring Intelligence Platform) is a platform that answers the question that bothers many senior managers in the AI era in light of the changes the labor market is undergoing: what is the best way to do the work.
The system provides every manager with three central resources for managing work: permanent employees who are the organizational foundation and carry institutional knowledge and organizational culture, freelancers and external experts, and artificial intelligence and automation systems. The system takes every process, project, or job and breaks them down into the building blocks that will allow the optimal answer for them in "AI time" while providing cost frameworks, recommendations for suitable external experts, and a breakdown of the best AI tools for the defined tasks.
"You feed into the system what you are looking for, and it builds the right mix of freelancers, full-time employees, and AI tools, and also says how much it will cost and how you can save," explains Steiner. The platform was developed with the help of Fiverr's recruitment team together with external freelancers and AI systems, and not by chance. A direct connection to Fiverr's market infrastructure is a built-in part of the platform, so that recruiting freelancers becomes a natural stage in workforce planning and not a point solution. Thanks to this, HIP serves as a decision-making tool that speaks both the language of the CHRO and the language of the CFO and creates significant cost savings while improving the way work is done at all."
Into this changing labor market enters Generation Z, which, according to Steiner, comes with a different perception of career, studies, and promotion. "They are a talented generation on levels that grew into a technological world. They really want to influence, to be part of a contribution, and they are not satisfied with one job. If in the past you would study, go to work, and retire from the same place, Generation Z today wants to do a role, and that is the transition to the next role. They have many tools, and I think they will invent amazing new roles. This is a generation that will relate less to titles and will be less interested in advancing to the management height but more across the board, and there will be less meaning for degrees and studies. They are autodidacts, and academia is not catching up with the gap. The subject of seniority and experience is very important, but there are roles that Generation Z will be able to bring.
Companies that succeed are those that will make the right and accurate architecture between human and machine, headcount and freelancers, and the previous generation and the next generation. This is the formula. Operational flexibility is the name of the game. I think that in the end, it is impossible without the human. A company that wants to live cannot do without the interpersonal meeting, the discourse, being together. Although this generation grew up with a computer, you still need diversity, everyone has a thought, it's not a template. The agent is a template because it's data, but different people can lead everything forward and will be more focused on leadership and coordination and will use tools."
Precisely after a hundred years in which the world of work tried to formulate rules, the current era is characterized by the fact that there is no new rule yet that will replace them. "If I have someone who studied in academia with 15 years of experience in a role, against someone who studied in academia or alone, and has zero years of experience but is an AI expert, I don't know who can bring more value to the company — the one who knows how to work with tools or the one who has experience. You need to combine, you need both, and there is no playbook for this, we are writing the playbook now, and it is intriguing where we will go."
