The revolt of AI orphans
Amazon: 14,000 people, Microsoft: 15,000, HSBC bank: 20,000. A few examples of recently announced workforce reductions directly attributable to AI. The World Economic Forum warns that corporate AI adoption does not merely eliminate jobs, but leads to the emergence of a new, unstable and frustrated social group. The “AI precariat” loses not only income and livelihood but is gradually pushed out of the economy, losing its social role. If this phenomenon becomes widespread, it may lead to political and social destabilisation, and in extreme cases even unrest and conflict.
AI increases social inequality at a systemic level: the technology replaces repetitive, rules-based work typically performed by the lower and lower-middle classes. As a result, a broad social class is pushed out of working life altogether, and the economy splits in two: a highly skilled, well-paid group and a low-paid, insecure mass. AI reinforces the emergence of a precariat in which more and more people live on vulnerable, project-based work rather than stable jobs. The result is a three-tier society: a winning, highly educated elite, an insecure precariat and a permanently marginalised underclass.
A structural break is occurring in the labour market: those who fall out cannot return because the new economy requires a higher level of knowledge for entry. This creates not simply masses of unemployed people, but an underemployed, underpaid group with no vision of the future. These people remain permanently on the margins of the system: they do occasional or low-value work but never return to the core of the productive economy.
Anthropic’s CEO warned that AI could eliminate half of entry-level white-collar jobs within a few years. According to the International Monetary Fund, 60% of jobs in advanced economies and 40% in emerging markets are already exposed to the impact of AI. The World Economic Forum’s 2025 Future of Jobs report shows that 41% of companies specifically plan workforce reductions due to AI by 2030.
Goldman Sachs analysed tasks in more than 900 occupations and concluded that around two-thirds of jobs in advanced economies are affected to some extent, with 25–50% of tasks in those jobs potentially automatable with generative AI. Overall, this means that work equivalent to as many as 300 million full-time jobs globally could soon be replaced by machines. The most affected areas are administration (46%), legal work (44%), and engineering and business analysis tasks, because they rely heavily on text and data processing. Not all jobs disappear; rather, parts of tasks disappear, so AI tends to complement work rather than replace it entirely. Even so, they estimate that around 18% of global labour demand can be automated, mainly in white-collar sectors, while physical jobs remain largely unaffected for now.
We face changes that call for a new, human-centred social contract. Striving to maintain employment at the highest possible level is not enough; individuals’ dignity and identity must also be protected. Universal mental healthcare is needed and will be a crucial element of future social stability. At the same time, rethinking traditional social security systems is inevitable, because the concept and role of work are changing fundamentally. We should become familiar with these concepts: basic income, community grants and guaranteed employment. These are not merely benefits, but important instruments of a new social balance if we are to avoid chaos. The permanent loss of employment opportunities is a real global risk: innovative initiatives such as “precariat labs” could allow governments, businesses and civil society organisations to experiment with new solutions together.
The concept of the precariat emerged in the late twentieth century, combining the words “precarious” and “proletariat”. French sociologists first used the term in the 1980s to describe workers in insecure employment. The concept became widely known through the work of British economist and sociologist Guy Standing and his book The Precariat: The New Dangerous Class (2011). Standing describes the precariat as a new, emerging social class characterised by insecure employment, income instability and a lack of social safety nets.
The “AI precariat” refers to the group pushed into insecurity and exclusion specifically as a result of artificial intelligence and automation. This group lives under unpredictable working conditions, without stable income or long-term security, trying to make a living from temporary, part-time or poorly protected jobs. Members of the precariat often lack employment-law protection and are easily replaced. This creates lasting insecurity of livelihood and social vulnerability. This group could become one of the greatest sources of social tension in the future.
This content and illustration were created through the synergy of the author’s human creativity, strategic decisions and intuition with AI’s capabilities for research, processing, organisation and writing.
