The Feminist Internet Research Network (FIRN) seeks to intentionally work towards feminist knowledge, shifting the ways we do technology research. With some of our network partners ready to share their findings, and others at the halfway point in the data analysis stage, we wonder: What's next? How does research play a role in advocacy spaces? How do we work towards the change we want to see by influencing policy? What is or should be a feminist approach to policy advocacy and policy shifts, specifically coming from the global South?
The open webinar “Policy reform: Working towards feminist transformation and change” was held on 22 June as part of the second FIRN (online) convening. A panel of speakers from different regions and backgrounds shared thoughts, experiences and stories around internet policy advocacy. They are: Mariana Valente (InternetLab, Brazil), Marwa Azelmat (APC’s Women Rights Programme), Helani Galpaya (LIRNEasia), Anita Gurumurthy (IT for Change) and Anriette Esterhuysen (chair of the United Nations' global Internet Governance Forum's Multistakeholder Advisory Group). The webinar was moderated by Chenai Chair (tech policy researcher).
We would like to especially recognise the work of our documenters, Liy Yusof and Sonaksha Iyengar, for making the textual and graphic records that allowed us to create this synthesis article.
A feminist approach to policy changes
FIRN has the long-term goal to inform and influence feminist activism and ICT policy making. Ours isn’t only a research agenda. It is a feminist knowledge-building strategy looking at global South realities to fill the gaps between research and policy making. What FIRN intentionally seeks is to go beyond research to impact on policy and advocacy.
Why is this important? As Helani Galpaya remarked in her talk, we need to impact policy capacities to build institutional transformations and change policy infrastructure: “One type of policy change is to change policy by improving and broadening policy capacities. That is, to change the people acting upon policy, the policy makers themselves, so you don't have to bang their door down 10 times saying ‘This is bad, this is not feminist,’ but they have awareness before they make policy, and this doesn't depend on civil society retrospectively acting. It's about building institutions, but that's hard.”
“One type of policy change is to change policy”.
It is important for advocacy to push policy changes to take place simultaneously at the national, regional and global level. As Anriette Esterhuysen pointed out: “There is a dynamic between decisions made at global level and holding governments accountable at national level. This links us to the power of collective action: feminist international movements to work together at a global level based on common goals, to try to achieve some agreements that reflect those goals at an international level and then work again at a national level to hold governments and other actors accountable to adhering to those agreements. The local versus global struggle is almost like a dialectical process, where you move from one to the other in a dynamic that is often uneven and with conflict, but that can be very powerful. […] Online GBV [gender-based violence] is a good example. Online GBV is a struggle fought at national and global level, but once there was global recognition of it being a real thing, a real issue, it became easier to take the struggle forward at national level.”
But what does a feminist approach to policy changes look like?
For Helani, the complexity of policy making around technologies, like artificial intelligence (AI) and machine-driven decisions, requires us “to go back to one of the core feminist principles, which is that process is as important as product. Both in research and policy making to find solutions. When governments implement large AI systems, we should be asking, Who’s present in the team? […] Who is creating the policy/system, for whose benefit?”
And as Anita Gurumurthy added later: “The question for feminism is not so much the power of machines, but the power of people who govern those machines. I’m worried about who machines give power to.”
Feminist influence in ICT policy and advocacy is also connected with our everyday activism practices. As Anriette wisely said at the end of her talk: “Something we can do as a form of self-strengthening and self-care is to be conscious about, at a human level, where we are in our lives as feminist activists. Is it right for us in this space and time to work at national level or global level? Lots of power and leverage in alternating between those levels and doing it strategically in terms of policy change window opportunities, but also in a way that makes sense in your own life.”
The role of research to influence policy spaces
How does research play a role to influence policy? What is important – evidence, stories, surveys, data findings, analysis?
Regarding the role of engaging with communities and movements in the work of policy change, Mariana Valente, based on the InternetLab experience in Brazil, commented: “In our experience, the work of organisations like ours that are working with research to do policy advocacy is much more effective when it's combined both with the work of social movements (and social movements meaning feminist collectives, activists, groups) but also the academia that is doing work specifically in different areas of gender equality and feminist research. Through those connections we can raise our sensitivity and have attention to issues that aren't always visible and seen. But also that connection makes us understand the links which need to be recognised so the different realities of groups of women are the nexus: first, interpreting of facts, second, localising contextual realities, and third, promoting narratives that are effective grounded and responsible. […] An example of this is a study we did on non-consensual sharing of intimate images (NCII). The mainstream media narrative was that it is hard to confront NCII because there is no specific criminal offence. Through case studies and activism in São Paulo, we found that the lack of a criminal offence wasn't the problem – all NCII cases depended on women having to hire lawyers. […] A new criminal offence type wouldn’t solve the problem as it seems to be on people's minds, and you wouldn't see that if you didn't look at the realities of women going through that, especially vulnerable women.”
“...the work of organisations like ours that are working with research to do policy advocacy is much more effective when it's combined both with the work of social movements but also the academia.”
But when data is being used to inform policy strategies, or moreover, when policy decisions are algorithm-driven, we must be aware that data biases can impact policy results in problematic ways. Feminist researchers and activists normally aren’t in the decision-making spaces to warn about the misuse of technology in policy making. “Often we’re not at the stage where artificial intelligence makes policy, but input into policy-making processes is given through AI,” Helani noted. “So the outputs that the algorithms and data spew out are highly biased, and you're making decisions about welfare without recognising there's such a huge group of problems.” That’s why she suggests, regarding the use of algorithms in policy making, “procurement rules calling for feminist audits, similar to environmental impact audits.”
On this subject, Anita Gurumurthy added: “While AI audit protocols are in development, we should work with examples of best practices from different parts of the world and what we can learn from them while we do the difficult but necessary task of contextualising them within our local social-structural frameworks. We need to encourage states to build capacities for AI audits, and advocate to build agreement, in the meantime, towards the need for and compliance with them. 
“Often we’re not at the stage where artificial intelligence makes policy, but input into policy-making processes is given through AI.”
The last point leads us to another relevant question: How do we address the growing power of private companies and big business?
Anriette highlighted that it is critically important “to avoid letting big internet companies establish their own regulatory regimes. We can use international systems and institutions to hold them accountable.” And she brought up an interesting point on alliance building: “South-South solidarity is traditionally important for us in the global South, but there's a dimension of South-North solidarity that's important and can be mobilised effectively, especially when trying to contain the power of big companies. We cannot even do that without working with activists in those countries where companies are registered.”
Anita offered us the example of big platforms’ accountability in content moderation: “Civility is now increasingly being enforced by platforms, not lawmakers or our conscience. What we want to do as feminists is recognise that content moderation is not merely a function of prohibited speech and identity.” To the question of the ethical problems of automated moderation versus human moderation, Anita answered: “AI is here to stay, and will need to be used to filter content because the scale of content is such that relying on human evaluators alone would be akin to not doing content moderation at all. The need for human intervention is of course an all-important issue. Context is necessary to understand hatefulness, and when language and images are used in conjunction and in ways AI can't yet process (perhaps never, to the extent of accuracy needed), for example, when sarcasm is deployed. The dangers for AI use arise when the deployment of AI is used to eschew explainability and accountability.
The power dimensions of the spaces we occupy
We also asked our guest speakers about the spaces and platforms for pushing for policy change around internet rights. Is it shifting from traditional spaces such as the UN Human Rights Council or even more recent multistakeholder spaces such as the Internet Governance Forum (IGF)?
With regard to influencing internet governance, for Marwa Azelmat, “gender-sensitive governance innovation is required to create an integrated framework at national and regional level. This needs and involves experimentation and international coordination, and engagement of a wide variety of stakeholders. Intersectionality should be embedded between sectors and stakeholder groups in addressing gender issues.” However, “while women are in the room at IGF meetings, when one moves through the various layers of leadership and responsibility, the number of women likely decreases. In addition to women's participation, we need to look at how receptive internet governance institutions are to women's issues. While participation is welcome, they're not expected to bring up women's issues. Old-fashioned ethos that all issues are gender-neutral, policies driven by trademark issues, or whatever the reason is, the issues of internet governance are not seen as gendered.”
Maybe that’s why feminists have been reluctant to get involved in internet governance. There is a sort of skepticism that can lead to a counterproductive strategy. But Marwa asked later: “Can we afford not to be part of the solution? If technology is to rewire many of the structures and practices that we hold dear, it is essential that we try to guide these changes, to ensure that the results support, rather than contradict, feminist visions.”
Nevertheless, often we are right to be skeptical. As Anriette described, in many policy spaces, “there is NO dialogue, NO debate, NO conversation. No one really listens to anyone. It is one prepared statement after another. It is polite, it is empty, and it feels powerless, even though the people who are speaking in it all have some kind of power.”
Perhaps we need to subvert and hack the characteristic "empty politeness" of international internet-related policy debate. For Anriette, hacking that empty politeness in a meaningful way requires us to be more disruptive and impolite, but also to bring to the table substantive content and analysis backed up by evidence. “Feminist activists need to have done the work of analysing specific intersections, and then describe them in a way that is understandable to people in other movements and disciplines and input this in writing into policy processes. This can combine well with insider and outsider strategies. If trying to influence processes from the inside (for example, by using available channels for comment, statements, participation, etc.) fails, one can use the same texts to challenge these processes from the outside and question their legitimacy.”
Perhaps we need to subvert and hack the characteristic "empty politeness" of international internet-related policy debate.
Anriette added: “I think it is important to remember that policy advocacy can take place anywhere – locally, in an institution, in a workplace, nationally, globally. Whether you believe it is worth investing in it depends on the context, the people and institutions. […] Don't make assumptions about where power is concentrated. Look at where power is concentrated on a certain issue. Look at your analysis and develop a strategy.”
The session closed with a reflection from Chenai Chair on how we are present in global policy and advocacy spaces: “It can seem when you engage in these global spaces that it's such a disembodied experience. I've been personally frustrated in these spaces, feeling like banging on a wall. But I'm grateful to spaces like these to move it beyond global level into local level. We need to identify who to work with, contextualise the power dimensions of the spaces we occupy, and decide where we choose to place our energies.”
. The Berkman Klein Center has published research which looks at the ways in which this is already happening. IT for Change's recommendations to the Indian Ministry of Electronics and Information Technology on content management also incorporate this necessary demand.