Financing caring economies beyond extraction: Feminist perspectives on tax justice, AI and development
•
Article summary
Notes from the Civil Society Policy Forum on 16 April, titled Financing caring economies beyond extraction: Feminist perspectives on tax justice, AI and development. As the IMF and World Bank deepen their engagement with artificial intelligence and digitisation as drivers of development, this session asks harder questions: whose labour powers these systems, who pays for them, and who is left out? Drawing on feminist political economy, panellists examined how AI and platform capitalism are reproducing colonial patterns of extraction, deepening inequalities, and further shifting unpaid care burdens onto women – and what a transformative alternative would look like.
Moderator
- Emma Burgisser, Economic Justice Lead, Christian Aid
Panellists
- Imene Cherif, Economic Justice and Research Advisor, MENA Fem Movement for Economic, Development and Ecological Justice (MENAFem)
- Amy McShane, Senior Gender Project Officer, The Bretton Woods Project (BWP)
- Dr. Lyla Latif, Chair, Committee on Fiscal Studies
- Daniela Behr, Economist, Women Busines and the Law, WBG
Organisers: Akina Mama wa Afrika (AMwA), The Bretton Woods Project (BWP), MENA Fem Movement For Economic, Development, and Ecological Justice, Christian Aid
A video recording of this event is available here.
Emma: We have been hearing a lot at these meetings about artificial intelligence and digitisation and how they can transform development, deliver jobs, and drive the global economy. But as feminists we are here to ask some deeper, more challenging questions – specifically about how AI’s rapid expansion risks reproducing colonial patterns of extraction, deepening inequalities, and further shifting unpaid care burdens to women. And before we begin, I want to acknowledge the many colleagues, allies and friends who could not be here – because of war, visa restrictions and funding cuts. The voices we cannot hear in this room are many, and that matters for this conversation above all others.
Lyla: The global financial architecture – the IMF, the World Bank, sovereign debt markets, OECD tax standard-setting, bilateral investment treaties – has consistently and structurally produced extraction: the movement of value from the Global South to the Global North, from the resource-rich to the capital-rich, from those whose labour built the system to those whose ownership of it the law has always protected. That logic did not emerge from nowhere. It is the same logic that produced colonialism, racial capitalism, and the patriarchal organisation of finance – the embedded assumption that women are not the proper owners of capital, do not control it, and cannot use it for enrichment. And it has found its way directly into international taxation. The international tax system was built on the principle that corporations are taxed where they are headquartered, not where they generate value. A company extracting minerals from the DRC or running a ride-hailing platform across Nairobi pays its principal tax in London, Dublin, or Delaware. Those rules were written in the 1920s by four economists from capital-exporting countries, at a time when most of Africa and Asia were under colonial administration and had no voice in the process.
The architecture of extraction is now being reproduced through digitalisation and deepened through artificial intelligence. Think of a domestic worker in Nairobi or Lagos employed through a platform app. She is matched, rated and priced by an algorithm. The platform monetises her reputation – five stars, three stars – but when the tax authority tries to locate her in the fiscal system to extend a social protection entitlement or a pension, the algorithm that governs her working life says she is an independent contractor. She has no employer. She pays no employment tax, receives no maternity protection or sick pay, and the platform that profits from every job she completes has no tax obligation in relation to her whatsoever. She is fiscally invisible – not because she is economically inactive but because the system was designed to see through her. And crucially, the AI being trained on her behavioural data – her routes, her ratings, her response times – is generating value captured by a corporation paying no meaningful tax in the country where she lives. She is subsidising, through her labour and her data, the very system that excludes her from fiscal recognition.
We need new vocabulary for this. Techno-feudalism – the idea that we have moved to a cloud-based order in which platform owners extract rents from populations with no exit – is useful. But for the Global South context I want to go further: what we are witnessing is techno-neocolonialism, where formal political independence coexists with substantive economic subordination reproduced through digital infrastructure. And beneath it operates what I would call algorithmic dispossession – the process by which platform design choices, tax classification systems, and digital financial infrastructure combine to make the labour, assets, and fiscal contributions of Global South populations invisible to the systems that are supposed to protect and serve them. Two things must be pursued together in response: a binding UN Framework Convention on International Tax Cooperation that establishes source taxation, automatic information exchange, and gender-responsive tax design as foundational obligations – not permissive options; and the genuine integration of care economy analysis into everything we are designing, from AI governance to development finance to tax reform.
Imene: In the platform economy, we are not just the product – we are also working for these platforms for free. When you post on social media, when you share content, when you interact with an AI system and make it smarter, you are creating value without compensation, without a contract, and without any social protection. Influencers invest their own equipment, their own time, their entire family lives – and they generate revenue for the platform while being classified as independent contractors with no employment relationship whatsoever.
Go deeper into the supply chain and the feudal logic becomes clearer. Data centres are the new land. They generate rent. And to function they consume enormous quantities of electricity and water – electricity that is prioritised over household consumption across the Global South, water that contributes to climate crises from Latin America to Africa. In Tunisia we are now seeing young women choosing delivery platform work precisely because of the freedom it appears to offer – riding motorcycles, setting their own hours. But they have no contracts, they bear the cost of equipment, they absorb all accident risk, and they are working in conditions that are technically illegal because there is no recognised employment relationship. The narrative of flexibility and freedom is a big lie. And in the skilled labour market, platforms are doing something even more insidious: fragmenting tasks to the minimum possible unit, putting them online, and using freelancers with no social protection, no maternity coverage, no health insurance – who also have to buy their own laptops and pay for their own internet. We are not against technological progress. What we need is regulation, and institutions like the IMF and World Bank have a role to play in requiring it.
Amy: I want to bridge what Iman has described with how the World Bank specifically is approaching this – and how its approaches can exacerbate rather than reverse the dynamics of extraction. The Bank’s current jobs agenda, with its focus on getting a billion people into work and promoting female labour force participation, is not a neutral project. Adding digitalisation and AI to that agenda is not a neutral add-on either. What we consistently see is that these systems rely on unpaid and underpaid care work mostly carried out by women – and the jobs women will be filling under digitalisation are precarious, flexibilised, and constructed as a positive when in reality they pass risk onto individuals who already face structural inequality. Women who carry disproportionate care responsibilities will be concentrated in the lower-paid, less secure, more precarious forms of digital work.
To contextualise the stakes: in 2018 the ILO estimated that 16.4 billion hours globally were devoted to unpaid care work every day. Valued at minimum wage, that constitutes 9 per cent of global GDP – $11 trillion – and women carry out three quarters of it. When the Bank talks about getting women into work, it has to address that barrier first. Otherwise we are just creating a triple burden: invisible care work, additional precarious paid work, and the personalised debt and economic shocks that come with it. The Bank has recognised care – there is the childcare window in IDA – but these are nowhere near sufficient responses to a structural problem of this scale. Solutions require abandoning austerity measures, financing essential public services, and addressing the tax and debt dynamics that the Bank has the power to influence.
Daniela: Legal frameworks are the architecture of economic life – they determine who can work, who can own, who can borrow, who can become an entrepreneur. And in too many places they do so in a way that systematically disadvantages women. Our most recent Women, Business and the Law report assessed the economic rights of women in 190 economies and found that women today hold only two-thirds of the legal rights of men. Not a single economy in the world – rich or poor, North or South – has achieved full legal equality for women. Half of economies do not explicitly prohibit gender-based discrimination in access to credit. Fifty-five economies still prohibit women from working in certain sectors – typically the sectors that pay better. Forty-three economies restrict inheritance rights for women. Removing these legal barriers could increase global GDP by around 20 per cent.
But the most important finding this year is that laws alone are not enough. Globally, laws supporting women’s economic participation are only half enforced. Weak institutional capacity, limited access to justice, and deep social norms mean that formal equality on paper produces very little in practice. And on technology: safety is an economic issue. No one can participate in the economy – online or offline, digital or analogue – while living in fear. AI and digital platforms are creating new domains of exposure to harassment and discrimination, and legal frameworks are not keeping pace. The good news is that progress is possible and is happening: in just two years, 68 economies enacted 113 legal reforms advancing gender equality, with most progress in sub-Saharan Africa. The question is how to accelerate it.
Questions and answers
Professor, Clark Atlanta University / CEO, Tido Tech International: We are facing an opportunity here, not just a threat. If young people – especially women – get involved in AI development at the early stage, there is real potential to reverse some of the biased algorithmic logic being embedded right now. The ethical design of these systems is not fixed. I am building an AI Climate Innovation Centre in Nigeria, and this conversation is making me more aware of what I have to be careful about in data gathering. How do we introduce ethics into the training of our youth – especially in rural areas where women are the primary agricultural producers?
Imene: Regulation is the answer, and ethics has to be baked into what regulation requires. But more than that, we need to insist that the platforms themselves are legally responsible – not just the governments that host them or the workers they misclassify.
World Bank public specialist, tax policy reform project in West Africa: My question is on the evidence for formalisation for women entrepreneurs. From a tax perspective, giving everyone a tax ID makes them legible – but is there evidence that the transition from informal to formal actually delivers benefits to women, rather than just exposing them to taxation without the corresponding services?
Daniela: The evidence is mixed, and you are right to interrogate the assumption. Legal reform can create a virtuous cycle, but only when enforcement and social norms shift alongside it. What we see clearly is that where discrimination is codified or tolerated – in access to credit, inheritance, sector restrictions – formalisation does not automatically unlock opportunity. The supportive frameworks matter as much as the formal legal status.
IMF research analyst and financial literacy content creator, Italy: Why are legal barriers to women’s finance so entrenched – is it cultural norms, or assumptions financial institutions make about women? And is there evidence of initiatives on the ground that are moving the needle?
Daniela: Both. Discriminatory laws send a signal – to employers, to the judiciary, to households – that discrimination is acceptable. That is why reform matters even before enforcement catches up. On financial institutions specifically: research shows that women pitching to venture capitalists are asked failure questions while men are asked growth questions. No legal framework can directly change that, but an environment that actively disincentivises discrimination shifts incentives over time. Italy actually does well on supportive frameworks for female entrepreneurship in our report, including some that explicitly cover digital entrepreneurship.
Emma: How do the findings translate into recommendations for regulation, and what conversations does your team have with colleagues providing day-to-day policy advice?
Daniela: We provide both a mirror – where countries currently stand – and a map – how to move forward. Operational colleagues working directly with ministries are using our data in policy dialogues. In Sierra Leone, findings from our report entered directly into discussions with the Ministry of Labour and contributed to reform; Sierra Leone was one of our top reformers. The Bank’s gender strategy aligns with the ten topics we cover. And I want to be direct: civil society creates the bottom-up pressure on governments without which these reforms do not happen. The Sierra Leone example would not have occurred without civil society engagement.
Amy: The ILO’s definition of decent work has four pillars – promoting jobs and enterprise, guaranteeing rights at work, extending social protection, and promoting social dialogue – and the Bank does technically have a partnership with the ILO. We are concerned that this is not translating into practice. Engaging with unions as key stakeholders, upholding international labour standards, abandoning austerity, and financing essential public services through progressive taxation and debt reform – these are the actual levers. And shareholders -including Global North governments – need to be willing to use their political capital to support these solutions. Without that, the gender language stays in strategy documents and never reaches programme design.
Imene: Two things give me hope. The UN Framework Convention on International Tax Cooperation is still being negotiated – this is a live process and civil society has a real voice in it. We should all be in those rooms. And regulation of AI and large platform corporations has to be on every institution’s agenda, including the IMF and World Bank. Ethics cannot be an afterthought – it has to be built into how these technologies are governed from the start.
