Critical Approaches to AI Policy: Recap of the 2026 Digital Politics Summer School by Cherry Cheuk Kwan Chiu
The intersection of artificial intelligence and public policy is moving faster than ever.
However, much of the mainstream debate around AI regulation remains focused on
technocratic fixes, corporate compliance, and western-centric models. In June 2026, the
Digital Politics Summer School, co-organised by the Center for Digital Ethics and Policy
(CDEP) at Loyola University Chicago and the Digital Society Research Group (DISC) at
Manchester Metropolitan University, convened to disrupt this narrative.
Bringing together a global cohort of academics, civil society advocates, and the first-ever
cohort of Critical AI Policy Virtual Fellows, the four-day event challenged participants to
critique existing AI frameworks through the lenses of power, justice, and human dignity.
Here is what went down at the 2026 school, capturing the vibrant energy and critical
takeaways from a week of radical reimagining.
Day 1: Policy at Play and Dismantling the AI Hype
The event kicked off with a welcome from the organising team, Jessica Elias, Florence
Chee, Daniel Joseph, and Adi Kuntsman, setting a collaborative, international tone bridging
the US and the UK. Dr. Florence Chee delivered the opening keynote, “Policy at Play:
Designing Critical AI Systems in Times of Polycrisis,” which used an unexpected but highly
relatable metaphor for AI: the global egg supply chain.
Chee argued that AI is a floating signifier that learned from a flawed, pre-existing human
condition rife with extraction, colonialism, and structural inequities. “Putting technology into
the hands of people isn’t enough if the system is out of reach,” Chee noted, pointing out the
historical failures of laptop-drop policies that lacked long-term infrastructural maintenance.
A dynamic afternoon panel featuring Kyla Williams Tate, Dr. André Brock, Dr. Momin M.
Malik, and anthropologist Dr. Todd Harple explored how models inherently reshape the
realities they purport to measure. Harple highlighted the constraints of corporate capitalism,
noting that “models are always built for people with the means to buy”. Brock pushed the
panel’s critical edge further by questioning the fundamental worldview built into these tools,
asserting that adding a few more seats to a broken policy table will not alter an extractive
trajectory. Instead, the panel closed with a call to shift AI toward “tools for conviviality”,
technologies designed to augment human capability and dignity rather than automate
everyday life out of existence.
Day 2: Accountability and Dignity in Long-Term Interactions
Wednesday pivoted toward presentations from the Virtual Fellows, structured around the
overarching theme of Accountability.
AI and Biological Fragility: Dr. Urooj Raja exposed the “virtual silence” surrounding the
environmental cost of AI at the federal policy level, urging regulators to target data centre
carbon footprints The Guardrails of Global Languages: Kate FitzGerald presented a multilingual investigation of chatbot safety features against conspiracy theories, revealing that low-resource
languages (like Ukrainian) are highly susceptible to linguistic inconsistencies and colonial
biases.
Platform Accountability & Misinformation: Lucie Čejková’s content analysis of major social
media community guidelines highlighted a glaring accountability gap: most platforms
explicitly avoid defining misinformation, and mentions of generative AI are virtually nonexistent.
AI Companionship: Mahla Alizadeh mapped out the relational risks of long-term emotional
engagements with AI voice assistants and companion bots, calling for UX designers and
policymakers to address “relational harms” that emerge dynamically over time.
The afternoon shifted to Dignity. Dr. Eva Nieto McAvoy introduced the concept of
posthumous custodianship, exploring the ethical minefield of digital necromancy, using AI to
reanimate the data traces of the deceased. She advocated for a dignity-led approach that
reframes data not as an asset to be monetised by platforms, but as a relationship requiring
ongoing care and community repair.
Dr. Malcolm Katrak rounded out the day by introducing a “labour logic” to AI governance. He
argued that corporate alignment initiatives (like Anthropic’s Claude Constitution) treat
fairness as a technical metric to optimise, resulting in “bias washing”. True ethical alignment
requires embedding labour protections across all layers, from micro-level workplace
surveillance to planetary compute infrastructures.
Day 3: Challenging the Periphery – Sustainability, Equity, and Trust
Day 3 brought a powerful shift in perspective, centring viewpoints from the Global South and
informational peripheries. Luis Rossi analysed the platformisation of AI in Latin America,
defining how foreign tech monopolies exert “infrastructural power”. Rossi offered an
optimistic and tangible blueprint to counter winner-take-all dynamics: the creation of a Latin
American counter-stack focused on local, open-source, offline interoperable AI models
tailored to the region’s pedagogical and linguistic realities.
Stephanie Kasaon (Kenya), Dr. Chikezie E. Uzuegbunam (South Africa), and Dr. Aristea
Fotopoulou (Greece) presented a brilliant comparative framework highlighting how
peripheral sites are used as unregulated regulatory sandboxes. Uzuegbunam shared a
cautionary tale: South Africa’s national AI policy framework had to be temporarily withdrawn
because the text itself relied on a western-trained AI model that fabricated several citations.
Fotopoulou exposed a similar trend in Greece, where a selective rollout of OpenAI’s
ChatGPT Edu in elite secondary schools risks widening pre-existing domestic digital divides
while creating profound technological dependency.
The collective remedy? Moving away from weak, unilateral local regulations and instead
anchoring national reforms in unified coalitions, such as the African Union’s Continental AI
Strategy, to generate true upstream accountability against global tech providers.
Day 4: Playing with Infrastructure via Creative Practice
The final day, led by Dr. Jessica Elias, shifted from diagnosing problems to co-creating
functioning futures. Elias made an impassioned case for creative practice (poetry, zines,
photography, and textiles) as a non-extractive, participatory research method.
Traditional policymaking takes community insights without offering valuable reciprocity.
Creative practice, by contrast, treats lived experience as genuine expertise and invites
policymakers into community spaces not as top-down authoritarians, but as learners. It uses
the social force of imagination to build the relationships and grassroots systems necessary
to make alternative futures endure.
The school closed with the announcement of the CDEP + DISC Digital Policy Observatory,
whose launch will be celebrated in Chicago this September, alongside the publication of
policy briefs and toolkits developed by the fellows.
Final Thoughts: The Road Ahead
As the summer school concluded, our chat boxes overflowed with gratitude for an
intellectually energising and generous week. While critical approaches to AI can often leave
us feeling defeated by the sheer scale of the polycrisis, the overriding sentiment of the 2026
cohort was one of determined optimism. By centering human dignity, honoring regional
contexts, and utilising creative practice, we can actively shift the trajectory of AI policy
toward solidarity, equity, and public accountability.
