Artificial Intelligence: Use of AI in the world of work Consequences and co-determination

dossier

"...The use of artificial intelligence is currently causing worries and fears among many employees, Schmitz continued. "Self-controlled processes are leading to a new division of labor between humans and machines, although this has not yet been sufficiently researched." Urgent questions of control, relief or burden, qualification and job security must be clarified. (…) "Artificial intelligence is a tool, a means to an end. AI applications must serve to improve the quality of work and quality of life." AI systems should be designed in such a way that employees' scope for action and design is expanded. "Activities must be enhanced through targeted qualifications. This can be supported by collective agreements and supported by new job profiles." ver.di press release from December 9, 2020 for the joint study by ver.di, IBM and the Federal Ministry of Labor, see the study and other articles on the topic:

• Subject to the algorithm: In the US healthcare system, nursing jobs are increasingly being assigned via platforms. Trade unions warn of a dramatic deterioration in working conditions
  "When it comes to the topic of artificial intelligence, people like to talk about how many professional groups will disappear as a result of its use and how many jobs will be cut as a result. Much less is discussed about the fact that the use of AI technologies is also an attack on the working conditions of wage earners. Because AI is used by capital to change the balance of power and make more profitable use of human labor. Here, it is not so much that people are being replaced by robots, but that wage earners are being subjected to an opaque, AI-supported algorithm Exploitation and exploitation processes are organized in the interests of capital and against wage earners. This can be observed in the USA. In recent years, a new sector has developed within the so-called platform economy, which has primarily set its sights on the healthcare system, according to the title of a study published a few months ago by the AI Now Institute, which critically examines the phenomenon US hospitals. The healthcare sector in the US is a growth industry, albeit fraught with contradictions. On the one hand, US society is aging rapidly. By 2040, one in five Americans will be over 65 years old During the lockdown phase of the Covid-19 pandemic, many US nurses resigned due to stress.The National Nurses United (NNU) union says that one in five nurses quit in the first few years of the 2020s because they no longer wanted to put up with undersupply and staffing shortages, or because they felt vulnerable to the dangers of the Covid-19 virus due to a lack of support from management. (…) This situation has brought US tech capital onto the scene, which senses strong growth opportunities. The idea: Vacancies should increasingly no longer be filled by permanent employees. Instead, hospitals can post their open shifts on online platforms. Job seekers then compete for these shifts, for example by posting their salary expectations on the platform. The person with the lowest salary expectation then gets the job. While wage earners compete on these application platforms for the lowest possible hourly wage, the platforms ensure that temporary nursing staff are disciplined. (…) There is no transparency for the employees. The platform determines the level of wages as it sees fit. In return, employees are attracted with the promise of professional flexibility. (…) However, upon closer inspection, the promises of more flexible forms of work turn out to be a nice illusion. In fact, the platforms create a fragmentation of work processes on site. Employees recruited through platforms do not know the teams with which they work their shifts. They are unfamiliar with specific, often informal work processes on site.This means that you can hardly build relationships of trust with colleagues - which means there are no opportunities for informal acquisition of new knowledge. However, all of these things are prerequisites for the company's collective self-organization of employees, which are thus placed in the way of obstacles. The hospitals in the USA, which operate largely for profit, can thus save costs and increase profit margins. The same applies to the platforms that are rewarded for their business model with rising prices on the stock exchanges. According to the AI ​​Now Institute, transnationally active speculative capital funds have long since bought into these platforms. They sense a lucrative business for their investors. (…) The NNU union fears the loss of hard-won rights for nursing staff, such as the right to a statutory minimum wage. Nursing staff recruited via platforms would also have to bear higher unemployment insurance costs, the union said. This would also make it more difficult for unions to organize. The Amazon group has now also entered the market, which makes the threat to workers' rights even more urgent. To date, U.S. unions have tried, largely in vain, to enforce long-term unionization at Amazon.” Article by Christian Bunke in the Junge Welt from September 24, 2026 
  • See the protests - most recently on August 27 "From Palo Alto to Washington: Nurses joined patients and supporters in imaginative acts of civil disobedience against Palantir" - in the dossierPalantir, Amazon & Co.: The digital deportation aides of the US immigration authority ICE
• The work is running out. Hurray, right?! The ILO report on global youth unemployment due to low growth and AI points out absurdities of technological development
  • 257 million young people without a secure income
    “According to an ILO report, low growth and artificial intelligence are leaving more and more young people behind worldwide. This particularly affects women
    According to a report by the International Labor Organization (ILO), global youth unemployment has risen again for the first time since the corona pandemic: to 12.4 percent. This means two million additional unemployed young people worldwide, out of a total of 67.3 million. The UN organization's labor market economists predict a further increase by 2027. The increasing NEET rate, i.e. the proportion of those who are neither working nor pursuing further education or training, is particularly worrying. This rate climbed from 19.7 percent to 20 percent in the past three years. This means that 257 million young people worldwide have virtually no prospect of a secure income from formal employment. (…) Even those who have a job often cannot make a living from it, let alone make a good living. According to the ILO, almost nine out of ten young workers in sub-Saharan Africa work in the informal sector - without contracts, regular wages or social security. Many of them move from one odd job to the next or run several small businesses at the same time. Because there are no formal security systems, young people receive little government support in the event of a crisis. Unemployment quickly becomes a threat to their existence, which is why those affected have to take on any job, no matter how precarious, in order to somehow survive. The situation is particularly dramatic for young women. According to the ILO, more than two thirds of all young people worldwide who are left behind in the labor market and education system are female. In regions such as North Africa and the Arab States, the NEET rate among women is over 40 percent.The experts primarily cite unpaid care work in the household as reasons, but also childcare and structural discrimination in the labor market. “We will only succeed in reducing these rates if economies specifically create jobs for young women and support them in balancing family and work,” demands Sukti Dasgupta, director of the ILO employment department. (…) In addition, the weak global economic momentum is slowing down job prospects. According to forecasts for 2026, global growth will fall to 3.1 percent, with young people primarily affected. (…) There are now also technological innovations such as artificial intelligence that are revolutionizing the job market, primarily in entry-level jobs in middle office and administrative areas. According to the ILO, 6.1 percent of youth jobs globally are severely affected by AI (…) According to the ILO, these gloomy prospects have long had political consequences. (…) Many young people feel forced to migrate. At the same time, there are increasing youth protests around the world. (…) Against this background, the ILO experts are calling for drastic reforms from governments, starting with ethical regulation of AI systems. “In addition, states must expand social safety nets for informal workers, invest specifically in social infrastructure such as affordable child care and strengthen the care economy in order to offer women the prospect of a secure income…” Article by Felix Sassmannshausen from August 12, 2026 in New Germany online above
  • Global Employment Trends for Youth 2026: Youth unemployment rises as young people face a harder road to decent work
    is the complete 78-page ILO studycan be found on the English-language ILO page 
  • The work is running out. Hurray, right?! The ILO report on global youth unemployment points to the absurdities of technological development
    "In many areas, artificial intelligence is taking over the dreary typing, sorting and managing of bullshit jobs. Life could become more beautiful: people would have more time to pursue their passions, more capacity to find joy in raising children, to follow their curiosity and to think about how to explore the infinite expanses of the universe sustainably and together. But there is this catch: Anyone who is wage-dependent in this society and wants to eat has to sell their labor. But if no new jobs are created, old ones are If they are no longer there, there is a problem. And that has long been a reality: the latest figures from the International Labor Organization (ILO) show that millions of young people around the world are left without any prospects - and the trend is increasing again. Women, who already often take on unpaid care work in the household, are hit particularly hard: Where formal jobs are disappearing, they have to make ends meet in the unregulated informal sector. The answer to this cannot be a conservative return to traditional family or gender images or the mantra-like liberal promises of "Lifelong learning." Demanding that young people optimize themselves even more and make themselves usable by hook or by crook, just to be one step faster than the algorithms they replace, further increases the psychological burden - and with it the longing for conservative myths that only simulate relief in the here and now. The ILO's reform proposals - from ethical regulation of AI to the expansion of social safety nets for informal workers to investments in affordable childcare – are necessary.But this raises the question of their financing, particularly in the Global South. And the ILO proposals only tamper with the symptoms, leaving the actual question untouched: How do we produce our social wealth and who receives it? Many young people seem to see the question more clearly. We can only hope that the frustration doesn't turn into resignation or even more intense punching and stabbing. But that it creates energy for a better future for all people on this planet.” Comment by Felix Sassmannshausen from August 12, 2026 in New Germany online 
• Economist warns of AI consequences: States are more likely to rely on repression than redistribution
  "A study by US economist Daron Acemoğlu calculates how governments would react to the threat of unrest caused by AI-related inequality. The result: at a certain level of automation, repression becomes more cost-effective for states than redistribution.
  True miracles are expected from the spread of artificial intelligence in the economy - at least if you take the stock market prices of AI companies as a benchmark. Estimates of the coming productivity gains run into the trillions. But increasing productivity could also mean that millions of jobs will be devalued or become unnecessary. “Let’s prepare for a jobs apocalypse,” advises the British Economist. Profits for companies, losses for wage earners – this increases inequality and endangers social peace. A new study by US economists outlines how a government would respond to a threatened revolt: with redistribution - or with repression? The results are disturbing. (…) Even if unemployment does not explode, the renowned US economist Daron Acemoğlu still expects a redistribution from bottom to top. Because AI makes work cheaper, to the benefit of companies that are in the hands of only a few people. “The inequality between labor and capital – the gap between the income generated by labor and the income generated by assets – will almost certainly increase,” Acemoğlu predicts. (…) There is therefore a fear of a revolt among the elite. (…) In a theoretical study, US economist Acemoğlu, together with Arda Gitmez and Mehdi Shadmehr, has now tried to outline the options for action for governments in this situation. In their model, automation increases inequality and therefore the likelihood of revolt. If it is successful, “the capital stock will be expropriated forever – a significant political risk that capitalists must take into account,” the study says.The “capitalist state” is therefore faced with the alternative of either redistributing the new wealth through taxation or simply repressing the population “and thus creating more scope for further automation”. (…) Unless the risk of an uprising is very low or the capital stock is not very low, the capitalist state prefers repression. A higher capital stock in turn promotes more automation and therefore more repression.” Article by Stephan Kaufmann from July 27, 2026 in the Frankfurter Rundschau online 
• Artificial Intelligence and the Working Class: Within companies, the biggest users and advocates of AI are found at the executive level
  “…We have been inundated with AI and the discourse surrounding it for years, and virtually all visible changes have been negative. Companies are replacing human customer service with inferior bots. Scammers are trying to pass off AI-generated art, music, videos and books as real. People fall in love with chatbots, fall into AI psychosis, or are driven to suicide by a “helpful” chatbot. Huge new data centers are being built against the wishes of local residents. And yet billionaires continue to praise AI and have recently even begun to chastise the general public for “not using it enough.” (…)
  All of this can lead to the question: Why are companies spending tens of billions - why is the global economy spending trillions - to develop a technology that almost no one (except scammers) seems convinced of?
  The reason: AI is not for you, but for your boss.
  Taken to its logical conclusion, the promise of AI is: “Every person is a CEO.” Instead of working yourself, you have a personal assistant who follows your instructions with utmost precision. He never contradicts you, never tells you that your idea is bad, and never demands more money or better working conditions.
  Within companies, the biggest adopters and advocates of AI can be found at the executive level.
  OneStudy cited by the Wall Street Journal June found that 40 percent of non-management employees said AI saved them no work time, while another 27 percent said it saved them less than two hours of work time per week. In contrast, 19 percent of top-level executives (CEOs, CFOs, etc.) said AI saves them more than 12 hours of work time per week, while another 57 percent said it saves them between 4 and 12 hours of work.
  Viewed from this perspective, it is also understandable that established technology companies such as Google and Amazon have invested so heavily in AI start-ups and that the biggest breakthroughs in generative AI have come in the area of ​​programming. Tech company CEOs are chasing a fantasy where physical capital can completely replace human capital and where they can be bosses without employees. (…)
  When AI is used in the workplace, the worker's role shifts from that of a direct generator of knowledge and creative content to that of a quasi-supervisor who must review everything generated by the machine. In the boss's dream, this worker's workload will double, then triple, and continue to increase as tasks are merged - although this remains part of the boss's fantasy for now. (…)The current threat to AI is not primarily that creative employees' royalties will be cut for work already done, but rather that it will become more difficult to find future work. Freelancers are much more affected by AI than permanent employees. In fields ranging from graphic design to stock photo production to copywriting, jobs have dried up, and much of the remaining work involves “correcting” poor AI first drafts. (…)
  The easiest response to this as an employee would be to reject the entire project in disgust. But even if the current AI bubble is likely to burst, the progress made will remain. Some of the world's worst people will continue to lead large corporations and continue their efforts to implement these advances - even if it takes another ten to twenty years. (…)
  AI is unlikely to make billionaires' sci-fi dreams come true. But just in case it can make some dreams come true, we should fight for our dreams instead.” EnglishArticle by Karl Zimmerman from July 10, 2026 in UE News the United Electrical, Radio & Machine Workers of America – Karl Zimmerman is research director at UE (“AI and the Working Class: The Economy, the Environment, and the Future of Work”, machine translated)
• [For many, AI is not an assistant. She's a boss.] Forget the "AI jobs apocalypse." The real danger of AI lies in the control and surveillance of workers
  “A new divide is emerging: between workers who use AI at work and those who are controlled by it
  The real threat that artificial intelligence poses to the world of work is not just job losses - it is the growing gap between those who use AI to expand their skills and those whose working lives are increasingly dominated by opaque, AI-powered surveillance and control systems.
  The debate about artificial intelligence and its impact on workers is heading in the wrong direction. On the one hand, there are warnings that machines threaten millions of jobs. On the other hand, it is claimed that AI will increase productivity. Both perspectives ignore what is already happening in workplaces around the world, from Britain to Kenya to the United States.
  For some, AI can help free daily work from routine tasks. These are often people in better paid positions with greater personal responsibility: analysts, consultants, lawyers, scientists, managers. In these professions, AI – provided it is used to support employees and not replace them – can act like a co-pilot. It can aid human judgment, speed up routine tasks and create space for more creative thinking.
  For many others, however, AI is not an assistant. She's a boss.
  It's used in scheduling and monitoring tools, route optimization software, and automated performance dashboards - all systems that decide who gets what shift, how long a task should take, and whether someone is performing to their maximum potential. In these workplaces, AI is not something you use. It is something that observes and controls you.
  This is the new divide we should all be paying attention to.
  A third of employers in the UK are already using “bossware” technology to monitor their employees’ online activities. This already widespread surveillance of employees provides a glimpse of what is to come. (…)
  And in the long term, the same methods of algorithmic management and monitoring currently being honed in warehouses, delivery trucks, and gig work platforms will likely expand to corporate headquarters, hospitals, and schools. This can already be seen at companies like Amazon, where software developers report being monitored and pressured to use AI to increase productivity - even if, paradoxically, it slows them down. And Meta plans to track and record its employees' keystrokes, mouse movements and clicks to train its AI models. Some of the same workers who are currently benefiting from the rise of AI are on the verge of losing that advantage at some point.
  My own research over the past decade on the coexistence of workers and AI, cited in the 2024 White House Economic Report, suggests that the most pressing issue surrounding AI's impact on the world of work is not immediate mass unemployment. Rather, it is the widening gap in skills, autonomy and well-being between those who are allowed to work with AI and those who must be managed by it. Many jobs will remain in the future, but they will be more pressured, more fragmented and less humane.
  This is important because work is not just about income. It is also about dignity, trust and self-determination…” EnglishArticle by Nazrul Islam from May 11, 2026 in The Guardian online (machine translated)
• Four-day week: The only answer to the AI ​​revolution
  “AI destroys jobs and monitors employees down to the last detail. A trade unionist has a clear demand – and it is hardly surprising. Artificial intelligence (AI) is changing the world of work. More and more companies are justifying workforce reductions with AI. (…)
  Microsoft Office is used almost everywhere in companies, be it through Microsoft 365 or Teams telephony. This creates a consistent digital data room for companies. The prerequisite is a database through which customer data and individual work steps are evaluated. Customer data and data from administrative employees are brought together in workflow management systems. This makes it easier for companies to put employees under pressure to conduct internal benchmarking. Programs such as Microsoft Copilot or Purview make it possible to evaluate the tone of voice, choice of words and emotions in telephone calls and online meetings.
  Google has also expanded its product range with AI. The AI ​​assistant Gemini is integrated into Google Workspace to intelligently support the use of specific applications. Using AI, the data is not only evaluated in isolation, but also linked together. Comprehensive profiles of employees' work methods and performance can be created from separate information such as communication data, customer contacts and work patterns.
  Reducing working hours as a counter concept
  These developments also concern trade unionists. Even if AI falls short of its “lofty promises,” it has negative consequences for employees,writes Vishal Reddy on labornotes.org.He asks how workers can ensure “that technological progress is redefined according to their standards and not those of the oligarchs”? His answer is the 32-hour week. (…)Reddy works for themAmerican campaign “WorkFour” , which advocates a four-day week with full wage compensation. In order to be able to compensate for the declining volume of work, at least in terms of business, a reduction in working hours with full wage compensation is the right answer.But the increasing pressure to perform due to new technology is also an argument for shortening working hours…”Contribution by Marcus Schwarzbach from May 5th, 2026 in Telepolis 
  • See also our dossier:  What hasn't worked so far in terms of collective agreements is now being tried out in individual companies: the 4-day week
• [Wave of layoffs in the USA] The big tech companies promised that AI would revolutionize the job market - just not like that
  “Oracle is reportedly laying off thousands of employees, joining an already long list of tech giants cutting staff while investing hundreds of billions of dollars in AI data centers. Microsoft laid off 15,000 employees last year. Amazon cut 16,000 jobs in January. Atlassian laid off 10% of its workforce as part of its AI focus. Block cut 40% of its workforce, arguing that AI could take over much of the basic programming work required. Meta, which has the express goal of developing godlike "superintelligent" AI, has reportedly laid off 700 employees while expanding a stock option program for a handful of top executives.
  Perhaps this shouldn't surprise us: Executives at big tech companies have long warned that AI would lead to job losses. You may have just forgotten to mention that these losses are not necessarily due to AI tools replacing human workers, but rather to the same old, boring business reasons as in the pre-AI era: If you spend too much, you usually end up having to cut costs…” EnglishAnalysis by Allison Morrow from April 1, 2026 at CNN (“Big Tech promised AI would disrupt labor — just not like this”, machine translated), see also:
  • The threat is growing: AI is responsible for 25 percent of layoffs
    “It is still difficult to measure the actual added value of AI tools. Nevertheless, the new technology seems to be costing more and more people their jobs. The tech industry in particular continues to be severely affected. (…) In the summer of 2025, an MIT study found that 95 percent of companies, some of which have invested heavily in AI projects, have not yet been able to derive any significant profit from them. Nevertheless, the effects on the labor market are already clearly noticeable. According to recent data from Challenger, Gray & Christmas, more than 52,000 tech jobs have been cut in the US so far this year - including 18,720 in the last month alone. Large corporations such as Meta and Oracle were affected. However, the report also points out that the workforce cuts at Dell in particular drove up the numbers, bringing the total to its highest level since 2023…”Article by Noëlle Bölling from April 5th, 2026 at t3n.de 
  • But see also:Tech workers against Big Tech? Political potential for labor disputes in the age of AI
• Struggles of the modern working world: With digitalized employment relationships, the negotiation of labor rights is also changing
  "Between 1960 and 1961, workers at a Fiat automobile factory in Turin, Italy, conducted a kind of field research. They interviewed workers on site and drew maps of their workplace in order to understand the production processes. Their result: Italy's leading automobile manufacturer was on the verge of massive labor unrest. They were right, in the summer of 1969 a series of mass strikes spread across northern Italy. They had their origins in Turin Fiat factories. “The workers’ survey at Fiat shows that information and analysis have always been at the center of workers’ organization, especially in the context of new technologies,” concludes the European Trade Union Confederation (EGM) in a guide on how trade unions deal with algorithms.
  Algorithms and the datasets they are based on are at the heart of moving manual, paper-based processes to digital platforms and tools. The so-called “algorithmic management” is the use of data-supported algorithms to transmit management instructions and to control work processes.
  According to the EGM, 79 percent of European companies now work with it. “With algorithmic control, work distribution and pay are determined in a way that is generally non-transparent and often discriminatory,” criticizes the EGM. If algorithms are used for wage models, for example, they are based on the discriminatory social status quo. This means that these conditions can be transferred to the digitalized world of work. In order to prevent this, explicit countermeasures must be taken through technical solutions and human review. In addition, “workers can be subjected to algorithmically designed punishments, including the loss of their jobs, sometimes without ever being able to communicate with a human supervisor,” fears the EGM. "In order to address these problems, it is essential that employees and unions negotiate the algorithm."
  A European model for dealing with algorithmic discrimination is the Hilfr cleaning platform in Denmark. There was a union agreement there to ensure corporate responsibility for algorithmic decisions. There are also signs that corporations are using explicitly collected data about workers to depress individual wages via algorithmic calculation models. This particularly affects the platform economy. This is the informal labor market in which temporary contracts are awarded flexibly and at short notice to job seekers, freelancers or part-time employees via online platforms. Classic examples are driving and delivery services. There are also attempts to reduce wages in this way in the German transportation industry. (…)
  If workers acquire the data, they can also be tools for industrial action from below. Using new technologies, workers can better understand corporations and their structures and check to what extent employers are complying with contracts. If a company is unwilling to negotiate a collective agreement, the insights gained from data analysis can be used for campaigns or legal proceedings. There are already some examples of this too. Uber drivers in Switzerland collected and analyzed their data to find out whether they should accept the salaries presented to them. An app was developed in Brazil that allows drivers to test whether offers from ride service platforms make economic sense for them. The basic principle of modern industrial action is that in the digital age it begins with understanding what data corporations have – and how they use it.” Article by Sarah Yolanda Koss from March 30, 2026 in New Germany online 
• Four strategies of the US unions in the fight for AI - one of them is the demand for a four-day week with 32 working hours as standard for everyone
  • Four strategies of US unions in the fight for AI
    “The current hype about artificial intelligence in business is causing a lot of fear among employees. According to a Reuters poll on AI, 71 percent of people in the US fear that “too many people will lose their jobs.”
    Wall Street and big tech are running a massive hype machine to justify their massive, risky investments in artificial intelligence, promising that it will provide a "productivity boost" - meaning fewer workers and more profits.
    However, employees can be reassured because so far it has mostly just been empty words. To date, artificial intelligence has hardly made any profits. It can be useful for a few tasks - such as creating rough drafts of computer code or summarizing huge amounts of data - but is otherwise rarely equivalent to human talent. (…)
    Artificial intelligence is a management power grab disguised as inevitable technical development. To combat this, workers can use four strategies that have proven effective in the past: identify the real problem; organize into unions; demand ransom; and block it. (…)
    The National Nurses United union published the last yearPrinciples of “AI Justice” , in which concrete dangers are highlighted, for example when an automated algorithm decides how many nurses should be scheduled for a shift or which examinations should be ordered for a patient. The union argues that computer systems cannot replace human expertise. (…)
    AI job losses may be a real threat in your industry, but not because automation can actually do your job well. Leaders may not care whether students are nurtured, real facts are reported, or patients are cured. They just want to make money. AI gives them an excuse to allow work quality to deteriorate.
    Software executive and critic Anil Dash recently noted that half a million tech industry workers have been laid off since ChatGPT's release, largely because executives "can now use AI as an excuse to fire employees they've long wanted to get rid of."
    Jobs for entry-level programmers have been massively cut, while experienced engineers are being retained to fix the faulty code that AI has produced. But where will the next generation of experienced engineers come from if they can't learn on the job as beginner programmers? These short-sighted savings give new bargaining power to experienced programmers, who could now drive employee-owned solutions to train the next generation. (…)
    When it comes to management's new AI and automation ventures, workers may refuse to let non-union contractors take the lead. An army of consultants has sprung up to advise employers on the introduction of AI in hospitals and schools - siphoning millions from the actual education and care sector. (…)
    In the next wave, many “hyperscale” data centers will be built that are ten times larger than those already built. The largest of these will use as much electricity as the entire city of Philadelphia.  The enormous labor demands of these projects give construction unions leverage if they use them: to recruit new members, reject contracts for projects that face the most local opposition, and demand concessions on public services and environmental protection.
    A wave of local grassroots initiatives prevented 25 data centers from closing last year. When unions join forces with citizens' groups, they can work together to get more from developers and governments, such as eliminating billion-dollar tax giveaways for data centers that can bankrupt local schools and the road network. In California, such coalitions have unionized gas and solar power plants and won some popular demands.
    At best, such “ransom” deals can increase the costs for management trying to push a new technology and buy time for workers to go on the offensive with organizing. (…)
    After all, the most degrading impact of AI is not just on our work, but also on our skills and imagination. When music and films are created by a robot patching together past works, it deprives both audiences and artists of new, bolder dreams.  Even in more routine tasks, we learn through practical experience. Artificial intelligence is not an unstoppable force of progress. If used the way CEOs want, it would actually dry up the source of progress: employee know-how.
    Resisting management's abuse of power in technology is an important step toward taking responsibility for the world we shape at work—and keeping the path to a better world open.” EnglishArticle by Keith Brower Brown  from March 5, 2026 in LaborNotes (“Four Union Strategies to Fight on A.I.”, machine translated)
  • The labor movement’s response to AI? Give us our time back
    “When it comes to artificial intelligence, there is a significant gap in objectives between business and workers.
    Should AI trigger a technological boom, companies intend to use it to wage all-out war on labor. They will use efficiency gains to cut jobs, monitor workers and worsen working conditions. Even if AI falls short of its lofty (perhaps inflated) valuations, the working class could still suffer serious consequences as our economy and retirement savings increasingly depend on investments in AI.
    Workers feel that companies have set a trap for them. A recent Pew survey found that most workers are concerned about AI in their workplace, with fear highest among low-wage workers. Only 17 percent of Americans believe AI will have a positive impact in the next 20 years. Despite this collective fear, the ten richest tech billionaires increased their combined wealth by half a trillion dollars last year, largely due to speculation in AI.
    And yet, in the face of these troubling developments, unions and allied politicians have largely agreed on defensive and procedural solutions.The organizing and negotiating principles of the AFL-CIO focus on defensive tools: transparency, consultation, procedural guarantees and staff development.
    Although these measures are necessary, they are not enough. Compared to the rapid development of AI, they are progressing too slowly and require consistent enforcement throughout the contract term. Even when they gain widespread support, they fail to generate the mass enthusiasm needed for collective action. These tools aim to preserve what workers still have, rather than fighting for what they could currently gain. They will not change who ultimately benefits from technological progress.
    So how can workers ensure that technological progress is redefined according to their standards and not those of the oligarchs?
    THE 32 HOUR WORKING WEEK
    We may not have control over the economy, but we can control our ambitions. The rise of artificial intelligence offers us a chance to set an economic agenda that gives people hope that technology will improve their lives, not worsen them.
    A four-day, 32-hour week with full pay and benefits should be at the heart of this aggressive agenda. The demand for shorter working hours shows how the benefits of technology can flow not only upward, but also outward. If the promise of AI is that it “saves us time,” then it can give us our time back from work. (…)
    AtWorkFour – the national campaign for this demand – we have identified four starting points that unions can use to achieve shorter work weeks. It is urgent to start this fight again now. Workers are already looking at AI with concern, which gives us a tailwind, but the extent of actual displacement is still limited - meaning our negotiating leverage remains intact. (…)
    PROMOTE MILITANCE
    There is a growing consensus across the labor movement that we need more militant, strike-ready unions; two indispensable prerequisites for this are solidarity and class consciousness. The negotiations over a four-day week with 32 working hours strengthen both.
    Every employee can imagine this immediately: more free time, more three-day weekends, more time for family and social life. A four-day week can bring workers together despite major differences in type of work and pay because everyone benefits from it. This demand also heightens class consciousness. On one side is an overworked, burnt-out working class; on the other, a class whose life is characterized by leisure. As Shawn Fain, president of the United Auto Workers union, put it, the real freeloaders in our economy are the “passive income masters” who profit from the work of others…” EnglishArticle by Vishal Reddy dated March 18, 2026 in LaborNotes (“Viewpoint: Labor’s Answer to A.I.? Give Us Our Time Back,” machine translated) – Vishal Reddy is a union negotiator for nursing home workers and executive director of WorkFour, a national campaign to make the four-day, 32-hour work week the standard for all workers.
• [Berkeley University study] AI in the workplace can result in more work and burnout 
  “A new study shows: More artificial intelligence in companies does not necessarily increase employee productivity, but rather condenses their work
  Artificial intelligence can significantly reduce the workload by, for example, relieving employees of simple routine tasks. Then there is more time for the core tasks and the creative work process. Many companies hope to achieve greater efficiency through the use of AI tools. But according to a recently published...Study from California The opposite often happens: artificial intelligence in the workplace does not reduce work - it intensifies it. (…) The two scientists Aruna Ranganathan and Xinqi Maggi Ye from the Haas School of Business at the University of Berkeley in California accompanied 200 employees of a US technology company for eight months. The result: After introducing the AI ​​tools, employees worked faster and longer, sometimes during breaks and in their free time. They also took on new areas of responsibility - and not because it was required of them. If AI tools take over simple routine administrative tasks, employees have more time for more demanding, creative and therefore more “valuable” activities. However, together with the often playfully simple operation of the AI ​​instruments, a new dynamic can develop in the workplace: you can quickly send a new task to the AI ​​while you take care of something else at the same time. The perceived increased productivity can then shift more and more to breaks or free time. (…) However, this pattern does not lead to better work performance in the long term and, according to the scientists, can end in states of mental exhaustion and burnout. This in turn is reflected in poorer work results and even employee illnesses. Industrial psychologist Florian Schweden from the Institute for Work Design and Organizational Development in Hamburg is also familiar with this dangerous process. (…) But according to Florian Schweden, AI can put pressure on employees even if it is not perceived as a helping partner, but as a competitor that threatens one's own job.In one of his studies, employees' fear of being replaced by AI even resulted in a case of "robomobbing": Employees of a logistics company placed wooden pallets in the way of the self-driving robots so that they would not work as efficiently. In doing so, they wanted to show “that people who can climb over something are still needed.” (…) However, industrial psychologist Florian Schweden does not want to demonize the use of AI. For him, it is important that companies, together with their employees, clearly distribute the roles of humans and machines before introducing AI tools…” Contribution by Frank Bäumer from February 22, 2026 at BR24
• Legal violations in the employment relationship: Who is liable for the use of AI in the job?
  "... The use of artificial intelligence (AI) has now become indispensable in many companies. The European Union has tried to keep up with this development and issued the regulation on AI in 2024 (AI regulation). However, many liability issues in particular remain unresolved after the EU Commission withdrew the draft of an AI liability directive. (…)
  In the event of damage caused by the aforementioned legal violations, the employer will primarily be held liable. When using AI systems, the operator, here the employer, who bears the entrepreneurial risk and is therefore responsible for the organization of his business and the work equipment, has, among other things, a duty of care to organize their use. This is especially true if they are operated and/or monitored by employees. This also illustrates the training obligation under Article 4 of the AI ​​Ordinance, according to which the operator must impart AI skills to its employees if they are involved in the operation and use of AI systems. But the employer can also be held liable apart from the above-mentioned duty of care, since the employer must accept breaches of duty and negligence on the part of the employee (analogous to Section 278 of the German Civil Code) insofar as the employer acts as a vicarious agent of the employer. This applies to all activities within the employer's scope of duties, even if they are carried out contrary to instructions. Only when an internal connection with an employer's obligation can no longer be determined will an attribution be made. (…) The employer who has been sued can, taking into account the principles of case law on internal compensation for damages (…), take recourse against the employee if the employee is responsible for a breach of duty. (…)
  With regard to liability for breaches of duty, the general liability privileges under labor law apply. While in the general law of obligations in the event of breaches of duty within the contractual relationship according to Section 280 Paragraph 1 Sentence 2 of the Civil Code (BGB), it is assumed that the debtor is responsible for the breach of duty, Section 619a BGB makes an exception to this for the employment relationship in favor of employees: In the event of breaches of duty by employees, it is not assumed that they are responsible for the breach of duty. The employer therefore bears the burden of explanation and proof for this. In addition, according to case law on internal compensation for damages, the employee is not liable for slight negligence and only proportionally for moderate negligence. As a rule, the employee is only fully liable in cases of gross negligence or intent (subject to exceptions for reasons of equity). The case law has derived this liability privilege from Section 254 of the German Civil Code (BGB), the standard on contributory negligence. It is based on the idea that the employer also shapes the liability risk of the employees through the organization of operational processes. It is also stated in the literature that the success of an employee's recourse depends on whether training has been carried out in accordance with the requirements of the AI ​​regulation.Regardless of this, it should be noted that full liability of the employee due to intent will not automatically apply according to the aforementioned principles if the employee has intentionally violated a training requirement, because full liability due to intent presupposes that the employee also intentionally caused the damage (and not just the breach of duty). (…)
  The training requirement under the AI ​​Ordinance therefore affects all liability relationships relating to employees. Compliance with them is therefore essential. The problem is that the AI ​​regulation does not provide any specific guidelines on how the training should be carried out…” Guest article by Dr. Nadine Kramer and Dominik Süß from January 13, 2026 at LTO
• New research challenges the common notion that AI inevitably destroys good jobs - apparently without questioning AI itself
  “New research shows: the use of AI can increase wages. But for this to happen, companies must meet certain requirements.
  Artificial intelligence (AI) is changing our lives. For some, it is the ultimate savior: an increase in efficiency in the world of work and life, a solution to climate change and a source of scientific breakthroughs in medicine. Others – like OpenAI CEO Sam Altman or Elon Musk – predict the end of the world and the complete takeover of human work. (…) But the effects of AI are more complex (…)
  Instead of dreaming of an AI takeover, we should rather think about the power and political influence of authoritarian “techbros”. Only under democratic control and organized as a collective good, according to Surplus columnist Aya Jaff, can AI benefit broader society. Following on from this point, we – researchers at the Digital Futures at Work Research Center and the Institute for the Future of Work – examined the question of who benefits financially from AI at work in our study recently published in the British Journal of Industrial Relations. Specifically: How does investing in and using AI impact wages? (…)
  Economists have been arguing for decades that digital technologies primarily benefit highly qualified people. The concept of skill-biased technological change states that computerization increases the demand for skilled work, replaces less complex work, and thus increases wage inequality. But our results suggest that AI offers the opportunity to break this pattern. When companies introduce AI, it’s not just academics who benefit. Especially among employees with lower educational qualifications, it is clear that those who regularly use AI in their everyday work earn significantly more. In other words, if employees with simple tasks integrate AI into their work, they can share in the productivity gains. This challenges the popular idea that AI inevitably destroys good jobs…” Article by Felix Schulz from November 23, 2025 in the business magazine Surplus (“Artificial intelligence in the world of work: Who benefits?”) – please note: The complete article requires a subscription to the free newsletter; theLink to the English-language study but is possible without a subscription
• Does work have a future? Challenges and recommendations for action when working with AI
  "Can employees be easily replaced by artificial intelligence (AI)? We now come across this question almost every day. Some celebrate AI as a revolution that makes work easier and increases productivity. Others warn of massive job losses. But how big is the change really - and how can it be designed in such a way that it benefits employees? Let's see what the current research says. (...) Overall, the studies show that the much-vaunted job destruction caused by AI has so far failed to materialize. Even what was hoped for There have been no leaps in productivity - at least so far. Most people agree that AI is finding its way into the world of work much more quickly than previous innovations. The question that recent studies are concerned with is no longer whether or which activities will be automated by AI, but rather what conditions are necessary to ensure productivity, growth, innovation and a fair distribution of benefits, or will the pace continue to increase because the technology allows it? Living conditions - or are a few technology giants skimming off the profits (...) International research - from the OECD, ILO, World Bank and UNDP - shows very clearly: AI does not automatically lead to economic growth and prosperity for everyone. The amount and distribution of AI profits depend crucially on the framework conditions: namely on how states invest in competition, education, the labor market and social security (...) AI will bring advantages for some and disadvantages for some.The task of politics is to prevent social imbalances, to ensure social cohesion, and to maintain and create good work. This includes the expansion of social safety nets, qualification programs and targeted support for job changes. Increases in productivity must also benefit employees: newly gained freedom must not lead to greater pressure, but should make work more meaningful - also because creative, communicative and social skills are the basis for innovation. Companies and regulators alike have a responsibility to ensure that AI is developed and applied equitably.” Article by Tünde Fülöp and David Walker from November 13, 2025 in the A&Wblog 
• The counterproductive class struggle: About the privileged aversion of the educated middle class to AI
  "AI is evil, AI is fascism - this attitude towards "artificial intelligence" seems to be particularly widespread in social science circles. Well, it is partly understandable that some in this class, which has not previously seen itself threatened by automation, feel questioned. Large language models (LLM) are seen exclusively as a threat by many thinkers and knowledge workers as well as creative people. In addition to the criticism of AI that is worth considering, there is predominantly a mixture of ignorance, inconsistency and conceit Observe and a lack of self-reflection about what privileged gestures and what reactionary traits result from the blanket rejection of a technology (…) The result of this cynical attitude: Due to the inability to use discursive shades of gray, the field is left to those against whom one supposedly wants to defend themselves - on BlueSky If social media had already existed at the time when it was automated away (…) The refusal of large parts of the intellectual class in this country to prevent civil society from having an influence: How can our societies use this technology profitably, given the shortage of personnel in education, administration and health? “Artificial intelligence” is certainly not a panacea. And it doesn’t automatically lead to a better world.But coming up with the idea of ​​handing the entire field back over to the IT companies, which primarily work in the interests of shareholders instead of the common good, is intellectual negligence.” Article in Lorenz' newsletter from November 12, 2025 
• Teams will soon be able to “tell”: digital surveillance is gradually establishing itself in everyday work
  "In a few weeks, Microsoft will be introducing a new Teams function that can detect whether employees are actually in the office. Critics warn: What is intended as a convenience function opens the door to new forms of monitoring. A manual click on "In the office" or "In the home office" was previously enough. From December onwards, this will no longer be necessary: ​​Microsoft Teams can then automatically detect whether an employee is actually sitting in the company building - based on the WLAN to which their laptop is connected. According to Microsoft, the function is intended to help facilitate collaboration in everyday hybrid work. However, critics see this as a new instrument of control. The old question takes on new urgency: How much surveillance is normal in the modern working world today? And how much is still acceptable? (…) According to the General Data Protection Regulation (GDPR), location recording is only permitted if it serves a clear purpose and the employees' consent is voluntary. In companies with a works council, an agreement is also necessary. But in practice, such formalities are often routinely checked off. Employers emphasize that the functions serve efficiency - but the boundaries of control are fluid. What happens if Teams automatically indicates that someone is out of the office in the future? Does this information really have no consequences? Labor lawyers view this critically. For the Cologne lawyer Jens Usebach from the law firm Jura.cc, the new feature in Microsoft Teams alone does not represent complete “all-round monitoring”, as it primarily records presence in the office or home office and does not monitor all the activities of all employees seamlessly and without reason.(…) “Nevertheless,” the labor lawyer added in an interview with ntv.de, “automatic location tracking expands the employer’s control options.” What is crucial, says Usebach, is how extensively the tool is used. “If the logging takes place permanently and without a specific reason, the employer is moving in a gray area.” Usebach advises employees to have a detailed explanation of what data is collected, how long it is stored and who has access to it. Transparency is the key to preventing abuse. (…)Usebach is calling for more precise legal regulations for this. Although existing law already provides a framework, many new technologies are only regulated in general terms. In 2024, a draft for an employee data protection law was presented that is intended to create clear limits for surveillance measures - including the location of employees. But the law has not yet come into force. Employee monitoring not only has legal but also psychological consequences. Studies show that constant control creates stress, distrust and lower job satisfaction - and therefore has the opposite effect of what employers expect - less motivation instead of more efficiency. Contribution by Diana Dittmer from November 9, 2025 on n-tv.de – see the first reports here below
• [ver.di] Employee survey in the ICT industry: Burdens caused by digitalization and AI are increasing - great importance of company co-determination
  "Strong work intensification and transformation characterize everyday working life; the stress is high and the trend is increasing. These are the key results of an employee survey in information and communication technology (ICT) companies carried out by ver.di, in which around 13,500 employees nationwide took part in the spring and summer of this year. The results of the survey, which is based on the DGB Good Work Index, also confirm the great importance of company co-determination and collective protection regulations. "Organizational changes and digital Change, including AI, causes worries and increases the demands on employees,” said Florian Haggenmiller, ver.di federal ICT group leader. More than a third of those surveyed (37.6 percent) were worried about their professional future. "This shows how important it is to design the transformation processes with the involvement of employees and to secure them through collective bargaining and company regulations. In addition, the need for action in further training and qualification is huge; after all, around a third of those surveyed did not feel sufficiently qualified to meet future requirements." Further results of the survey are: Almost 71 percent of those surveyed currently see no or only limited opportunities for advancement. More than four fifths (81.7 percent) stated that organizational changes such as restructuring, outsourcing or reorganization often take place in their company. Around two thirds (64.6 percent) experience that the requirements for knowledge and skills are increasing as a result of digital change.Rushed work and time pressure are an everyday reality for the majority of those surveyed (53.3 percent), 48 percent report a high level of work intensification within the last year. The most common reason for work stress is too many simultaneous projects and processes; Around two thirds (66.1 percent) stated this. The results of the survey also show that meaning and collegiality are high. More than three quarters of those surveyed (77.3 percent) identify to a high or very high degree with their work. Collective cohesion is also rated well; Despite difficult conditions, almost nine out of ten employees (88.4 percent) report high or very high levels of peer support. The importance of company co-determination is also confirmed; More than 19 out of 20 respondents (95.5 percent) consider works councils to be very important.” ver.di press release from October 12, 2025 
• AI scrap instead of a productivity boom from AI tools in the workplace: When inferior “worklops” generated by AI need to be corrected
  “AI junk is taking over jobs. Workers said they view their colleagues who submitted low-quality AI work as “less creative, capable and reliable than they were before receiving the output.”
  A joint study by researchers at Stanford University and a workplace performance consulting firm published in the Harvard Business Review describes the plight of workers who must correct the "worklops" generated by their colleagues through AI, which they describe as work content that "masquerades as good work but lacks the substance to meaningfully advance a given task." The research, based on a survey of 1,150 workers, is the latest analysis to suggest that the introduction of AI tools in the workplace has not led to a magical productivity boom, but instead has only increased the amount of time workers say they spend correcting low-quality AI-generated “work.”
  TheHarvard Business Review study appeared the day after oneFinancial Times analysis of hundreds of earnings reports and shareholder meeting minutes filed by S&P 500 companies. It found that large companies have difficulty articulating the specific benefits of widespread adoption of AI, but have no trouble explaining the risks and downsides of the technology to their companies: "The largest publicly traded U.S. companies talk about artificial intelligence all the time. But beyond the 'fear of missing out,' few seem able to describe how the technology is changing their business for the better