Top Cases
Should operators of high-risk AI systems be legally required to report incidents and near-misses to a public registry?
The EU AI Act obliges providers and deployers of high-risk AI systems to log serious incidents and report them to national supervisory authorities, yet no equivalent obligation exists to disclose those incidents in a centralized, publicly accessible registry. When an AI-assisted diagnostic system misclassifies a scan, when a credit-scoring algorithm makes systematically biased decisions, or when a predictive-policing tool flags entire neighbourhoods without evidence, the affected individuals and the broader public currently have no reliable mechanism to learn that these failures occurred — let alone how frequently they occur across the industry. Proponents of mandatory public incident reporting argue that transparency is the foundation of accountability. A centralized registry would allow researchers to identify systemic failure patterns across vendors, enable civil society to scrutinize algorithmic harm, and empower regulators with cross-sector data that cannot be obtained from isolated, confidential reports. Aviation and nuclear sectors demonstrate that mandatory incident reporting — even when anonymous — produces measurable safety improvements over time. Critics counter that full public disclosure may chill incident reporting itself: if organizations fear reputational damage or litigation from every disclosed near-miss, they may invest more in concealment than in mitigation. They argue that mandatory confidential reporting to regulators, with anonymized aggregate summaries, preserves the safety signal without creating perverse incentives. Competitive sensitivity and operational security further complicate full public disclosure in healthcare and financial services. The question is whether public transparency is a precondition for meaningful accountability in high-risk AI, or whether regulator-only reporting with controlled public summaries achieves the same safety outcomes at lower systemic cost.
Should internet platforms be required to provide archival access to cultural heritage institutions before shutdown?
When a digital platform shuts down — a social network, a fan community, a creative writing host, a forum — it often takes with it years or decades of irreplaceable human expression. GeoCities, Vine, Google+, early Flash game portals, and countless forums have vanished permanently, taking with them millions of conversations, artworks, and social histories that cannot be reconstructed. Yet most jurisdictions impose no advance notice obligations on platforms, and cultural heritage institutions — national libraries, digital archives, university collections — rarely have the legal standing or technical access to preserve this material before it disappears. Proponents of mandatory archival access argue that digital platforms increasingly constitute the primary repositories of contemporary cultural memory. When they disappear without warning, entire communities lose their history. Requiring platforms above a certain scale to notify accredited heritage institutions at least 90 days before shutdown — and to provide a structured data export pathway — would allow for meaningful preservation without prohibiting the shutdown itself. Precedents exist in financial and telecoms regulation, where institutions must give advance notice of service discontinuation. Critics raise legitimate concerns: the cost and complexity of compliance for smaller platforms, privacy obligations conflicting with bulk archiving, and questions about which institutions qualify as accredited recipients. Some argue that the internet archive and voluntary crawling services already perform this function adequately, and that mandatory regimes introduce bureaucratic overhead without proportionate benefit. There are also genuine tensions between preserving the historical record and honoring deletion requests from users who expected their content to disappear with the platform. The question is whether cultural preservation obligations should attach to digital platforms, or whether the existing ecosystem of voluntary archiving efforts is sufficient to safeguard our collective digital memory.
Should AI companies require free, prior, and informed consent from Indigenous communities before training language models on endangered language data?
As large language model development accelerates, AI companies are increasingly scraping text and audio corpora from across the internet — including digitised recordings, transcribed oral histories, and community-published materials in endangered Indigenous languages. Many of these resources were created by and for specific communities, often with the explicit purpose of cultural preservation, not commercial AI development. The principle of Free, Prior and Informed Consent (FPIC), enshrined in the United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP), holds that Indigenous peoples must be meaningfully consulted and must give consent before projects affecting their resources or cultural heritage proceed. But current AI training pipelines largely treat language data as a public commons: if it is accessible online, it is assumed to be fair use. Critics of this approach argue that scraped Indigenous language data enables corporations to build commercial products — voice assistants, translation services, generative AI — that profit from cultural knowledge without returning any benefit to the communities whose elders, teachers, and learners created it. They contend that training on this data without consent violates data sovereignty, may distort or decontextualise sacred or restricted knowledge, and displaces community-led language technology efforts. Proponents of the status quo argue that mandatory consent requirements are operationally complex to implement at scale, that freely published material was implicitly made available for broad use, and that AI tools trained on Indigenous languages could accelerate revitalisation efforts. The question is whether the default should shift: should FPIC be a legal prerequisite for using endangered language data in AI training, or does requiring it impose barriers that ultimately slow the development of tools communities themselves may want?
Should cloud providers be legally required to guarantee API stability and data portability under open standards?
When a major cloud provider deprecates an API or silently changes its behavior, users and businesses built on top of that infrastructure can lose access to their own data overnight — not through any fault of their own, but because of unilateral decisions made by a vendor whose incentives do not align with user continuity. This is not a hypothetical edge case. It happens regularly, and the costs fall disproportionately on smaller organizations and individuals who lack the negotiating leverage or resources to maintain redundant infrastructure. User sovereignty in digital systems requires more than legal ownership of data — it requires practical portability. Without it, ownership is theoretical. Regulatory frameworks like the EU Data Act take steps in this direction, mandating switching rights for cloud customers, but enforcement mechanisms remain weak and coverage is uneven. Voluntary industry initiatives like the Cloud Native Computing Foundation promote open standards, yet adoption across proprietary platforms is inconsistent and often performative. Proponents of binding legal requirements argue that the market has had decades to self-correct and has not. Vendor lock-in is not an accident — it is a deliberate economic strategy, and users pay for it in reduced agency and higher long-term costs. Legally mandating that cloud providers expose data in open, interoperable formats and maintain API stability windows would rebalance power toward users without prohibiting innovation. Critics contend that imposing standardization mandates on a fast-moving industry risks freezing innovation, creates unequal compliance burdens between large incumbents and smaller providers, and that the technical complexity of defining open standards across heterogeneous infrastructure makes enforcement nearly impossible in practice. They argue that certification programs and procurement requirements offer a more adaptive path. The question is whether user sovereignty in cloud infrastructure requires the force of law, or whether market and community pressure can achieve the same outcome.
New Cases
Should internet platforms be required to provide archival access to cultural heritage institutions before shutdown?
When a digital platform shuts down — a social network, a fan community, a creative writing host, a forum — it often takes with it years or decades of irreplaceable human expression. GeoCities, Vine, Google+, early Flash game portals, and countless forums have vanished permanently, taking with them millions of conversations, artworks, and social histories that cannot be reconstructed. Yet most jurisdictions impose no advance notice obligations on platforms, and cultural heritage institutions — national libraries, digital archives, university collections — rarely have the legal standing or technical access to preserve this material before it disappears. Proponents of mandatory archival access argue that digital platforms increasingly constitute the primary repositories of contemporary cultural memory. When they disappear without warning, entire communities lose their history. Requiring platforms above a certain scale to notify accredited heritage institutions at least 90 days before shutdown — and to provide a structured data export pathway — would allow for meaningful preservation without prohibiting the shutdown itself. Precedents exist in financial and telecoms regulation, where institutions must give advance notice of service discontinuation. Critics raise legitimate concerns: the cost and complexity of compliance for smaller platforms, privacy obligations conflicting with bulk archiving, and questions about which institutions qualify as accredited recipients. Some argue that the internet archive and voluntary crawling services already perform this function adequately, and that mandatory regimes introduce bureaucratic overhead without proportionate benefit. There are also genuine tensions between preserving the historical record and honoring deletion requests from users who expected their content to disappear with the platform. The question is whether cultural preservation obligations should attach to digital platforms, or whether the existing ecosystem of voluntary archiving efforts is sufficient to safeguard our collective digital memory.
Should UK high street fashion retailers be legally required to stock adaptive clothing lines?
Adaptive clothing — garments designed for people with physical disabilities, chronic conditions, or motor impairments — remains largely absent from mainstream UK high street fashion. Currently, disabled consumers are often forced to rely on specialist suppliers at higher prices and with far less choice. Some advocates argue that mandatory inclusion in standard retail ranges would transform access to dignified, affordable fashion for the 14.6 million disabled people in the UK. Others contend that legislation would be heavy-handed, and that market incentives, consumer awareness campaigns, and voluntary retailer commitments are more sustainable drivers of change. This trial asks: should UK law compel high street fashion retailers to include adaptive clothing lines as a standard part of their product offering?
Verdicts
Show allShould security researchers receive legal immunity when responsibly disclosing vulnerabilities found through unauthorized system access?
Researchers who discover vulnerabilities in good faith and follow coordinated disclosure protocols d...
Should healthcare AI diagnostic tools require demographic bias audits before regulatory approval?
No decision could be reached based on the available votes.
Should climate fiction authors have an ethical obligation to offer hope rather than despair?
Climate fiction must not sanitize grief or loss; honest depictions of collapse and injustice serve e...
Should fashion brands be legally required to offer adaptive clothing lines?
Legal requirements push resistant brands to act, ensuring disabled consumers have equal access to cl...