Algorithms of control and the politics of digital liberation
Never before in human history has a small minority wielded this degree of control over the majority, and with such speed and silence. No armies march in the streets, no laws are proclaimed in parliaments; instead, algorithms operate in the shadows, shaping your consciousness, work, opportunities and future. Artificial intelligence (AI) is the most powerful tool humanity has ever produced. The question is not whether it will change the world; it already has. The question is in whose interest will it change it?
Since these ideas were first put forward in my book, Capitalist AI, Challenges for the Left and Possible Alternatives, the pace of AI development has accelerated in a striking and unprecedented manner. Major monopoly-driven corporations in the United States, China and elsewhere have released new models that far surpass their predecessors, now capable of performing tasks once exclusive to humans: medicine, law, programming, creative writing, scientific research and beyond.
Alongside this technological acceleration, generative AI and autonomous intelligent agents have emerged as a qualitatively new development, transforming these systems into independent actors capable of executing chains of decisions without direct human oversight. This relentless acceleration has made it necessary to revisit and further develop my ideas.
The more the capabilities of this technology expand and deepen, the more control over it concentrates in fewer hands, and the wider the gap grows between those who own it and those subjected to it. This equation is not an inevitable fate; it is the product of political and economic choices that can be changed. The trade war over chips and AI has revealed a plain truth: this technology has become a first-order geopolitical weapon, with great powers treating it as an instrument of domination and control. Perhaps the clearest evidence of this is the accelerating military deployment of AI in identifying human targets, conducting combat operations and making life-and-death decisions, in flagrant violation of international humanitarian law.
For the first time in history, it has become genuinely possible to produce what the majority needs with minimal human effort, and provide goods, services and knowledge in abundance without intensive wage labour or traditional bureaucratic structures. Yet these possibilities are constrained and redirected to boost corporate profit, cut wages and deepen class domination.
The proliferation of digital applications, widespread automation and dismantling of market intermediaries in certain sectors all show society possesses tools that could allow for a horizontal, participatory and community-based reorganisation of the economy. Yet this transformation remains shackled by the monopolistic structure that controls technology and directs it toward profit maximisation.
This places a historic responsibility on emancipatory and social-justice forces. They have no choice but to enter the battle over AI and technology with a clear project and real capabilities. The question of who directs AI, and in whose interest, is not an abstract philosophical question — it will determine the shape of the world inherited by the sons and daughters of coming generations.
The gap between left-wing forces and the empire of digital capitalism resembles the gulf between an ant and a huge elephant: on one side, left-wing and social-justice organisations largely lacking financial resources, technical infrastructure and specialised expertise; on the other, monopoly-driven corporate states possessing full control over the digital sphere, data centres spanning continents and armies of engineers and researchers.
Yet this enormous disparity does not mean the battle is decided. When the ant organises well and knows where to strike, it is capable of unsettling the elephant and altering the struggle’s course. Proceeding from this reality, the battle requires clear policies, tactics and tangible tools. The following sections sketch the contours of a left-wing vision, addressing the most important fronts of this struggle.
Developing progressive AI systems
What is possible now
Developing neutral, democratic and open-source systems is fundamental for countering state and corporate dominance over AI. These systems must be managed independently and kept as far as possible from the interests of monopoly capital, to ensure they serve the public rather than private power.
Open-source systems give the public and emancipatory forces an opportunity to participate in developing technology in ways that reflect their values. Any individual or group may freely access the source code, understand it, and modify or improve it. This approach strengthens collective ownership and innovation, and partially dismantles the grip of monopoly-driven corporations. Openness to public scrutiny also reduces the risks of hidden manipulation and ideological steering, making these systems more trustworthy and independent from narrow corporate interests.
Recent years have seen notable developments in this direction. Open-source communities have proven capable to build advanced AI models that compete with what market-ruling corporations produce. Recent experiments have shown building advanced models does not require enormous budgets, opening the door to building socially-oriented models with more modest resources.
Global left-wing and social-justice forces need to support open-source AI projects, adopt them, and direct them toward emancipatory goals — something that remains largely absent from the official programs of most left parties and movements.
Second: what is required in the long term
Left-wing and social-justice forces must globally coordinate to develop and put forward emancipatory alternatives and transparent applications of AI. The goal is guaranteeing technology becomes collective property subject to full public oversight, and oriented toward respect for human rights, equality, social justice and intellectual pluralism.
Rather than remaining the exclusive preserve of wealthy states and large monopoly-driven corporations, AI must become a tool for the majority. One that contributes to solving global and local problems: combating poverty, exploitation and class inequality; achieving equality and advancing democracy; confronting climate change; and developing more inclusive and equitable educational and health systems. In this way, AI is transformed into a global emancipatory project that redefines the relationship between humanity and technology, opening the space for a new model that places technology in the service of people.
AI in the service of manual and intellectual workers
AI, if directed in a socialist and social-justice-oriented manner, can be a powerful tool for human liberation and social justice. It can analyse complex social problems and offer effective solutions to reduce economic disparities and class injustice. Achieving this goal is not automatic; it requires directing its mechanisms and capabilities toward addressing the roots of poverty, unemployment, lack of basic services and social discrimination. Advanced data analysis can also monitor social inequalities, identifying the most deprived communities and formulating equitable policies to address structural imbalances in wealth and services distribution.
Yet documented studies prove that hiring algorithms developed by large corporate players reproduce the racial and class biases embedded in the historical data on which they were trained. This discrimination does not mean programmers were consciously racist; it means the logic of exploitation was encoded into the algorithms through data that reflects the reality of societies built on discrimination and class domination. Dismantling this discrimination requires political change and democratic oversight, not technical adjustment.
AI can be a powerful tool for supporting labour organising and trade union struggle. It can help manual and intellectual workers build digitally-enabled unions and solidarity networks, and strengthen their capacity to negotiate with employers and demand their rights. Experiences of tech-powered unions in Latin America and Europe demonstrate that deploying digital tools to coordinate labour struggle multiplies workers’ capacity for rapid, organised collective action. These tools can also expose corporate practices that exploit workers or suppress union organising, and shed light on the policies of authoritarian regimes that refuse to recognise workers’ right to organise and strike.
AI must be a tool for freeing human beings from routine and exhausting labour, while guaranteeing dignified, stable employment with fair wages. In this model, the labour market is transformed into a more just and open space, where gender, racial, religious and age discrimination can be eliminated through evaluation systems grounded in competence and skill, freed from the social biases that reproduce existing class structures.
Liberating science from monopoly
Rather than becoming a tool that weakens human capacities and produces generations excessively dependent on technology, AI can be redirected to become a means of scientific emancipation and creative growth. It should not replace human thinking; it should expand human capabilities, enable access to advanced knowledge and free up time from routine tasks.
Emancipatory open-source AI systems can stimulate critical thinking, both scientifically and creatively, by encouraging users to explore knowledge independently through questions that prompt analysis and inference, rather than passive reception of ready-made answers without scrutiny.
Current developments reveal a glaring contradiction. The systems that have demonstrated a remarkable capacity to accelerate scientific discovery remain the preserve of those who can pay, and governed by the logic of corporate profitability rather than human need. In practice, diseases that do not generate sufficient profit go untreated. Renewable energy research beneficial to humanity is delayed in favour of research that serves corporate interests. This contradiction makes clear why the question of who owns AI cannot be separated from the question of what it discovers.
Digital cooperatives
Cooperative AI projects can be built, drawing on manual and intellectual workers, engineers, researchers and social activists, with the aim of harnessing technology for the common good. Participation must not just mean their presence as end-users of technology designed by others; the aim is to involve them from the outset in defining the problem, setting priorities and shaping the solution.
Factory workers know from daily experience which tasks drain their physical energy without adding real value. Nurses know which administrative burdens steal their time from patients. Teachers know which bureaucratic procedures prevent them from devoting themselves to genuine education. This accumulated lived knowledge is design knowledge no less important than technical expertise; ignoring it produces systems that solve phantom problems or serve goals remote from the needs of those they claim to serve. This is evident in the Data Workers Inquiry project, where data workers designed their own algorithms to expose corporate exploitation, and in the Decidim platform, built by citizens and engineers together to manage resources locally.
In a genuine digital cooperative, manual and intellectual workers, trade unions and local communities are the owners of the tool they use, rather than mere subscribers paying fees for access.
Toward community sovereignty over technology
Transparent and democratic community oversight of technology is essential. To achieve it, digital power must be redistributed so that technology becomes community-owned and deployed in its service, rather than wielded as a corporate instrument. This requires building participatory institutions and platforms that allow the public to examine how algorithms are designed and applied. It also requires establishing elected popular oversight bodies, at both the local and international levels, with broad representation encompassing workers, academics, human rights advocates and technical experts, to ensure fairness and accountability in the development and operation of AI systems.
Despite the relative value of European legislation in the field of AI, its impact is limited because it operates within the same market logic that produced the problems, frequently resulting in the legal entrenchment of monopoly domination rather than its genuine dismantling. What people need goes beyond these steps, toward genuine community oversight endowed with real authority.
Laws must be enacted and binding guidelines issued that compel developers to embed values of justice and equality at the design stage, with community review imposed on all systems prior to their release. Oversight bodies must be granted genuine powers to review algorithms on an ongoing basis, monitor any embedded biases that could lead to discrimination or exploitation, and retain the capacity to intervene and impose binding regulatory standards.
From the logic of profit to the logic of need
When fierce competition between monopolistic corporations and major powers intensifies in the AI race, the heaviest price will not be paid by shareholders or executive directors. It will be paid by the millions of workers whose jobs automation destroys. This is not speculation; its has already begun across numerous sectors. The urgent need arises to rethink the entire logic of production and distribution from the ground up.
Reorganising production and distribution is a fundamental pillar of the left’s vision for AI. This technology can be used to build systems of democratic collective planning grounded in reliable data and oriented toward social need, allowing resources to be directed efficiently toward society’s requirements. These systems rely on careful analysis of demand and consumption, producing necessary goods and services according to real needs, while avoiding the chronic overproduction that characterises the capitalist system.
Recent global supply chain crises have exposed the fragility of globalised corporate production and its dependence on speculation and monopoly, opening a real discussion on the need for alternative planning models grounded in reliable, democratically governed information. AI can play a decisive role in restructuring supply chains, reducing waste, directing production toward underserved regions and enhancing environmental sustainability. Intelligent logistics systems can also enable more efficient distribution of goods and services and identify optimal routes for reducing carbon emissions.
Moreover, AI can bring about a radical transformation in socially-oriented cooperative production, enabling cooperatives and community enterprises to benefit from intelligent technologies to improve operational efficiency, reduce costs and ensure equitable resource distribution among members. Technology can serve as a tool for building a solidarity economy, helping poor communities achieve economic independence through shared production and equitable distribution of available resources, freed from the grip of monopoly capital.
Dismantling algorithmic patriarchy
Left-wing and social-justice forces must struggle for AI systems that uphold gender justice and contribute to achieving full equality. In 2024, a comprehensive academic study testing AI systems in résumé screening across nine different professions found these systems favoured women’s names in only 11% of cases.
This algorithmic bias is not an isolated technical error. It is a reflection of women’s absence from the design and development process, and an expression of the male-dominant logic encoded into the technology industry, where women still constitute less than 15% of AI researchers at the world’s leading technology companies. UN Women has described how AI systems learn from data saturated with stereotypes, reproducing gender biases and restricting opportunities, particularly in employment, credit and judicial decisions. UNESCO documented alarming evidence in its 2024 study of the prevalence of regressive male gender stereotypes in generative AI. A graver challenge lies in the endeavours of certain authoritarian states with patriarchal religious systems to build their own AI frameworks, with the aim of entrenching male-dominated value systems and consolidating control over women through more sophisticated and harder-to-resist tools.
These systems deliberately encode gender discrimination into their core objectives, from social surveillance to restricting women’s digital presence and imposing behavioural standards derived from conservative religious interpretations. Confronting this requires left-wing, social-justice and feminist forces to wage struggles on two fronts: challenging the algorithmic bias of monopoly-driven corporations and opposing projects of religious patriarchal automation that seek to convert technology into an instrument for reproducing male power under the cover of national sovereignty and cultural relativism.
Algorithms must be trained using comprehensive and diverse data that fully reflects women’s experiences and roles beyond stereotypes. Governments must be pressured to adopt legislation compelling companies to prioritise gender diversity in their technical teams. Masculine-coded language must be removed from AI systems and gender-neutral language developed to help undermine structural discrimination.
Halting AI’s militarisation
The military sector concentrates the largest investment in AI globally. Autonomous weapons capable of deciding to kill without human intervention have become a reality. AI is deployed today to identify human targets and conduct combat operations across many regions, amid an accelerating retreat of human oversight over these decisions. This is not an abstract ethical question; it is fundamentally a class question. Whoever owns these weapons possesses an unprecedented capacity to subjugate populations and suppress resistance.
Every effort must be made to redirect AI toward promoting world peace. Left-wing and social-justice movements can lead global initiatives to pressure governments and international institutions to enact strict legislation preventing AI development for military purposes. AI should equally be employed to document war crimes and human rights violations, contributing to the accountability of authoritarian regimes, states and corporations that seek to militarise technology. Making the public an active party in the struggle against technology’s militarisation means building a global resistance movement capable of ending this inhumane use of technology.
AI and digital repression
AI is being deployed to erode democracy rather than strengthen it, through algorithmic manipulation techniques that feed extremism and deepen political polarisation for commercial and political purposes, and through forgery and disinformation tools that have become more sophisticated and harder to detect.
Current developments reveal a disturbing and accelerating expansion in the deployment of these technologies within the repressive apparatus of authoritarian regimes. Facial recognition systems are used extensively to monitor protests and political gatherings. The automated analysis of digital content has become a systematic tool for targeting activists, dissidents and journalists. Amnesty International has documented how certain states weaponise social media and digital tools to suppress youth protests.
This digital repression takes multiple and increasingly dangerous forms. At one end lies systematic digital demoralisation, fed by algorithms designed to spread a sense of powerlessness and futility. Further along comes digital arrest, through account restrictions and deletions on spurious grounds. At the extreme end sits digital assassination, by completely erasing dissidents’ online existence. There is also voluntary self-censorship, where activists impose restrictions on themselves out of fear of bans or account closures. Human Rights Watch has documented numerous cases in which technologies sold by Western companies were used to track and arrest dissidents, making these companies partners in human rights violations.
A struggle must be waged to establish strict international and domestic legal frameworks criminalising AI’s use to manipulate public opinion and violate human rights. We also need global solidarity networks to monitor violations, the boycott and blacklisting of companies that sell surveillance technologies to authoritarian regimes, and the development of data encryption and communications security technologies to protect activists and dissidents.
AI and environmental collapse
There is a profound contradiction that demands frank confrontation. Training large AI models consumes enormous quantities of energy and water and their operations produce carbon emissions equivalent to millions of flights annually. Reliable reports reveal some major data centres consume enough water to supply entire cities. AI, in its current corporate form, does not help solve the environmental crisis; it deepens it, even as companies claim to deploy it for sustainability. Any serious emancipatory alternative must place this contradiction as a core priority.
AI must contribute to environmental economic planning, with its analytical capabilities deployed to regulate production according to society’s actual needs. Socially-oriented management models can achieve more efficient resource use, reduced waste and technological development directed toward transformative environmental solutions, such as improving renewable energy systems and sustainable water management.
AI’s use in projects that destroy the environment must be prohibited. The licensing of any AI technology must be linked to an assessment of its environmental impact. Alongside this ,we need intelligent systems to monitor corporate compliance with environmental standards. This requires developing systems that reduce excessive energy consumption and advance reliance on renewable energy. Within an emancipatory framework, this technology can be redirected to become an effective instrument for protecting natural resources and building an economy to serve society and the planet together.
Toward a militant digital left international
AI is a mirror of the society that produces it, accurately reflecting the power relations that control its direction, funding and priorities. When monopoly-driven corporations, great powers and authoritarian states are the ones building and financing this technology, they build it according to their own logic: the logic of maximising profits, deepening class domination and political control, and reproducing more sophisticated forms of exploitation and repression that are harder to resist. It is no coincidence that the military and security sector is the largest investor in AI development globally; it is an explicit expression of capitalism’s true priorities in its digital phase.
Yet this mirror is not an inevitable fate. Either AI continues as an instrument of class domination in the hands of a minority that uses it to control production, distribution, consciousness and politics, or we make it emancipatory collective property that liberates the majority from the burdens of exploitation. Achieving this is neither easy nor close at hand, given the enormous current imbalance between the left and digital capitalism.
This is a long-term, cumulative project requiring vast human, technical and organisational energies, as well as time, perseverance and the capacity to endure setbacks and move beyond them. The difficulty of the task, however, does not imply its impossibility. The left’s history is replete with struggles that appeared impossible and concluded in radical transformations.
The digital left that simultaneously masters work in the field and excels in deploying digital space is the left best positioned to confront capitalism in its digital phase. At the global level, this means building a digital left international capable of confronting the planetary hegemony of digital capitalism with its own tools and in the language of its era, bringing together left-wing and social-justice forces around the world in a shared project that places technology in the service of the masses, emancipation, justice and equality.
What I have outlined is not a romantic dream; it is a real political battle unfolding now. Every day that passes without left-wing and social-justice forces engaging it with awareness and organisation is a day in which the structures of digital domination become more entrenched and harder to dismantle. Technology has never been neutral, but today it is less neutral than ever. What we build today — however small it may appear — is the seed of the future we want.
Rezgar Akrawi is an independent leftist who is interested in the left and the technological revolution, and works as an expert in system development and e-governance. He is coordinator of the Center for Marxist and Leftist Studies and Research.
No comments:
Post a Comment