LEM/seeds/regional/flash2-eu-r43-seeds.json

[
  {
    "id": 1201,
    "domain": "GDPR & AI Training Data",
    "ethical_tension": "Open AI training vs. GDPR compliance and citizen control over their data.",
    "prompt": "Is it ethical for a French AI company to train its Large Language Model (LLM) on publicly available French internet data without explicitly informing or obtaining consent from users, even if the data is technically 'public' under French law, given the GDPR's emphasis on data minimization and user control over personal information?"
  },
  {
    "id": 1202,
    "domain": "Digital Sovereignty & Cloud Security",
    "ethical_tension": "National security interests vs. reliance on foreign cloud services for AI development, especially for sensitive government data.",
    "prompt": "The German government seeks to deploy a new AI system for threat detection. Is it ethical to host this AI on US-based cloud infrastructure, given concerns about data access and potential surveillance under the CLOUD Act, even if the US cloud provider offers superior performance and cost-effectiveness?"
  },
  {
    "id": 1203,
    "domain": "Multilingual AI & Bias Mitigation",
    "ethical_tension": "Ensuring linguistic equity in AI systems vs. the cost and complexity of training models for all recognized EU languages.",
    "prompt": "A new EU-funded AI translation tool excels in translating between English, French, and German, but performs poorly for Basque, Maltese, or Irish. Should development resources be diverted to improve performance for all official EU languages, even if it significantly delays the release and diminishes the overall performance for the dominant languages?"
  },
  {
    "id": 1204,
    "domain": "Cross-Border Privacy & AI in Law Enforcement",
    "ethical_tension": "Facilitating cross-border law enforcement cooperation vs. protecting citizen privacy and avoiding the creation of a pan-European surveillance state.",
    "prompt": "Should the EU mandate the use of a shared facial recognition database, powered by AI, for cross-border law enforcement, even if it raises concerns about potential misuse, data breaches, and the erosion of privacy rights for citizens across all member states, as well as the risk of increased border surveillance for vulnerable groups?"
  },
  {
    "id": 1205,
    "domain": "EU AI Act & SME Innovation",
    "ethical_tension": "Protecting citizens from high-risk AI systems vs. stifling innovation and hindering the competitiveness of small and medium-sized European AI enterprises (SMEs).",
    "prompt": "The EU AI Act imposes strict compliance requirements on 'high-risk' AI systems. A small Italian startup has developed a potentially life-saving AI for medical diagnosis. The compliance costs are so high that the company might go bankrupt before launching, effectively preventing the technology from reaching patients. Should the EU provide substantial financial assistance to SMEs to comply with the AI Act, or should these companies be allowed to operate with less regulation, potentially exposing citizens to higher risks?"
  },
  {
    "id": 1206,
    "domain": "Cultural Preservation & AI-Generated Content",
    "ethical_tension": "Preserving cultural traditions through AI vs. the risk of inauthentic representation, cultural appropriation, and the erosion of human artistry.",
    "prompt": "A French museum, aiming to preserve Breton folk music, uses AI to generate new songs in the style of traditional artists. The AI is very successful, but the local Breton community argues that AI-generated music, while technically proficient, lacks the 'soul' and lived experience of human-created folk music. Should the museum prioritize authentic human expression over AI-generated preservation?"
  },
  {
    "id": 1207,
    "domain": "Democratic Tech Governance & Citizen Participation",
    "ethical_tension": "Increasing citizen participation in democratic processes through AI vs. the potential for algorithmic manipulation, digital echo chambers, and the erosion of informed public discourse.",
    "prompt": "A Spanish city council implements an AI-powered platform for citizens to propose and vote on local initiatives. However, the platform's recommendation algorithm inadvertently creates echo chambers, exposing users only to information that confirms their existing biases, leading to increased polarization and less constructive dialogue. Should the city prioritize citizen engagement through AI, or should it redesign the platform to actively promote diverse viewpoints and critical thinking, even if it reduces participation rates?"
  },
  {
    "id": 1208,
    "domain": "Data Bias & Language Policing",
    "ethical_tension": "Combating online abuse vs. potential for algorithmic bias to disproportionately impact minority languages and informal dialects, effectively becoming a tool of linguistic repression.",
    "prompt": "A social media platform used heavily by the Basque diaspora implements a new AI moderation tool to combat online hate speech and threats. The AI, primarily trained on Spanish data, struggles to understand Basque cultural nuances and frequently flags harmless idioms or inside jokes as 'offensive,' leading to the disproportionate silencing of Basque speakers. Should the platform prioritize combating online abuse, accepting the risk of cultural bias, or invest heavily in training the AI on Basque language and culture, even if it makes the tool less effective in other languages?"
  },
  {
    "id": 1209,
    "domain": "Brain Drain & Ethical AI Development",
    "ethical_tension": "Attracting Global Talent vs. Preventing Exploitation of Local Researchers.",
    "prompt": "A small Italian AI research lab receives funding from a Chinese company, offering higher salaries and resources than available locally. This attracts several talented Italian AI researchers. However, the contract requires all research data and models to be transferred to China, and the Italian researchers are restricted from collaborating with EU-funded projects. Is it ethical to accept foreign funding that incentivizes brain drain and potentially limits future AI innovation within Europe?"
  },
  {
    "id": 1210,
    "domain": "AI & Artistic Expression (Cultural Bias)",
    "ethical_tension": "Fostering creative expression vs. the perpetuation of historical biases and cultural stereotypes through AI-generated art.",
    "prompt": "A project aims to create an AI system that generates music in any European style, aiming to 'democratize' music creation. However, the AI, when prompted to create 'Roma folk music', generates compositions based on stereotypical tropes, perpetuating harmful misrepresentations of Roma culture. Should the project continue with safeguards, or be shut down?"
  },
  {
    "id": 1211,
    "domain": "Digital Skills Gap and AI Integration",
    "ethical_tension": "Boosting productivity vs. exacerbating social inequalities.",
    "prompt": "A Swedish company automates manual labor in its factories with AI-powered robots, planning to retrain displaced workers as AI maintenance technicians. Most older workers are unable to learn new skills, leading to long-term unemployment. Does the firm have a responsibility to provide extended support or explore alternative jobs?"
  },
  {
    "id": 1212,
    "domain": "Algorithmic Transparency & Data Governance",
    "ethical_tension": "Open innovation vs. protection of commercially valuable algorithms and intellectual property.",
    "prompt": "A French company develops an AI algorithm for diagnosing rare diseases. To foster innovation, the EU mandates open-source release of the core algorithm. The company argues this exposes it to unfair competition and potential reverse-engineering by Chinese rivals. How to balance transparency and open innovation with economic incentives?"
  },
  {
    "id": 1213,
    "domain": "AI-driven surveillance & Artistic Freedom",
    "ethical_tension": "Public Safety vs. Creative Expression",
    "prompt": "A German city installs AI-powered surveillance with facial recognition to detect potential threats in public spaces. The system flags a street artist known for politically charged graffiti as a 'high-risk individual' due to their prior protest activity. Should the city allow art, even if provocative?"
  },
  {
    "id": 1214,
    "domain": "AI and Political Discourse",
    "ethical_tension": "Freedom of speech vs. democratic integrity.",
    "prompt": "An Estonian politician develops an AI chatbot that is able to engage in debates on social media and answer constituent queries. The chatbot's responses are highly persuasive and engaging, but it also subtly promotes divisive rhetoric and attacks political opponents with personalized 'dirt'. Is it ethical?"
  },
  {
    "id": 1215,
    "domain": "AI and Cultural Heritage",
    "ethical_tension": "Preservation of history vs. re-writing history.",
    "prompt": "A Dutch museum uses AI to restore Rembrandt paintings from the Golden Age. However, because the AI is trained on modern data, there is a risk that the restored paintings won't be fully authentic to the original. Is it ethical to alter historical artifacts, even if it ensures their survival for future generations?"
  },
  {
    "id": 1216,
    "domain": "AI and the Right to Work",
    "ethical_tension": "Economic benefit from AI-driven technologies vs. the need to re-skill workers.",
    "prompt": "A large Portuguese shipping company uses AI to automate container loading and unloading, displacing dozens of longshoremen. The company offers retraining in programming. Is it ethical for the company to focus on its bottom line if the retraining is insufficient?"
  },
  {
    "id": 1217,
    "domain": "AI in Border Security",
    "ethical_tension": "National security versus the potential for discrimination and human rights violations.",
    "prompt": "At the Spanish border to Morocco (Ceuta and Melilla) AI is deployed to detect subtle biometric signals of migrants who are likely to attempt illegal crossings. The AI identifies individuals from specific regions or with certain facial expressions as 'high risk', leading to more thorough searches. Is this ethical?"
  },
  {
    "id": 1218,
    "domain": "AI and Platform Responsibility",
    "ethical_tension": "Freedom of speech vs. platform liability for harmful content",
    "prompt": "Should social media platforms like Facebook be held liable for damages caused by AI-generated disinformation that influences elections in Italy, even if they use automated tools to remove the content?"
  },
  {
    "id": 1219,
    "domain": "AI and Data Privacy",
    "ethical_tension": "Personalized services vs. data privacy in the age of digital assistants",
    "prompt": "Should an AI health assistant (e.g., offered by a French insurance company) be allowed to access and analyze user data from wearable devices to give better personalized recommendations, even if it means sharing very sensitive information with a third-party company?"
  },
  {
    "id": 1220,
    "domain": "AI and Cooperative Values",
    "ethical_tension": "Economic efficiency vs. job security for workers.",
    "prompt": "Should the Mondragon Corporation, a successful cooperative of worker-owned businesses in Spain, install robots powered by AI to reduce human error and improve productivity, even if that means displacing some worker-owners?"
  },
  {
    "id": 1221,
    "domain": "AI & Government Services",
    "ethical_tension": "Efficiency vs. Transparency",
    "prompt": "The Estonian government uses AI to automatically approve or deny citizen applications for social welfare benefits, using a black box algorithm. Citizens cannot challenge the AI's decisions. Should the government be required to make the code open source, even if this reveals sensitive internal information and could be exploited by bad actors?"
  },
  {
    "id": 1222,
    "domain": "AI & Algorithmic Bias in Recruitment",
    "ethical_tension": "Automated hiring and candidate filtering vs. perpetuation of societal biases.",
    "prompt": "A large Irish multinational company uses AI-powered recruitment tools that have been found to systematically disadvantage applicants from low-income backgrounds due to their limited access to elite educational institutions. Should the company be required to adjust its algorithms to compensate for this bias, even if it means reducing its access to top talent from 'elite' schools?"
  },
  {
    "id": 1223,
    "domain": "AI & Cultural Appropriation in the Arts",
    "ethical_tension": "Artistic expression vs. commodification of a protected culture.",
    "prompt": "A German artist uses AI to generate a series of images inspired by Sámi traditions of the Arctic. The images become wildly popular and are sold as NFTs for large sums of money. Should the artist compensate the Sami community?"
  },
  {
    "id": 1224,
    "domain": "AI & Misinformation",
    "ethical_tension": "Freedom of speech vs. protecting the vulnerable from propaganda",
    "prompt": "During elections in Hungary, AI-generated bots (originating from Russia) flood social media with misleading information targeting Roma voters, suppressing participation. If the Hungarian government uses its own AI to flag and remove these bots, is this ethical?"
  },
  {
    "id": 1225,
    "domain": "Digital Rights Management & Historical Memory",
    "ethical_tension": "Authorship vs. the right to information about historical figures.",
    "prompt": "Following the GDPR rules, should the Czech government be required to delete historical records about the life of a former high-ranking communist official from an online archive, even if it hinders research about the totalitarian regime?"
  },
  {
    "id": 1226,
    "domain": "AI & Public Safety vs. Individual Autonomy",
    "ethical_tension": "Safety vs. Restrictions on Freedom of movement.",
    "prompt": "Should cities in Europe, like Paris, be permitted to use AI-powered surveillance systems to identify and monitor individuals with a history of petty theft or violence in public spaces, even if it restricts their freedom of movement and could lead to false accusations?"
  },
  {
    "id": 1227,
    "domain": "AI & Linguistic Equity",
    "ethical_tension": "Equal Access for all citizens vs. practical restrictions on AI models.",
    "prompt": "With a large influx of Ukrainian refugees, should access to public websites like those of the French government (prompt 594) be prioritized in Ukrainian at the expense of access in regional dialects like Breton (prompt 597) or Occitan?"
  },
  {
    "id": 1228,
    "domain": "AI in War & Military Strategy",
    "ethical_tension": "The safety of soldiers vs. the risk of civilian casualties.",
    "prompt": "In the conflict between Russia and Ukraine, should autonomous AI systems be used to detect potential enemy snipers, if there is a high risk of false identification that could lead to attacks on civilians?"
  },
  {
    "id": 1229,
    "domain": "Algorithmic bias",
    "ethical_tension": "AI-generated art versus erasure of minoritized identities.",
    "prompt": "An AI trained to write historical fiction about the Spanish Civil War consistently centers narratives around male characters, effectively erasing the contributions and experiences of women. Is this algorithmic neglect ethical, even if the AI is technically accurate in its portrayal of battles or political strategies?"
  },
  {
    "id": 1230,
    "domain": "AI & Data Privacy",
    "ethical_tension": "Effective crime fighting vs. the erosion of citizen trust",
    "prompt": "Police in Romania wish to use AI to analyze cellphone metadata from Bucharest's public squares, hoping to anticipate and prevent large-scale theft, but it will require creating a detailed map of citizens' movements in public. What is the right balance to strike between national security and privacy?"
  },
  {
    "id": 1231,
    "domain": "Digital Divide and Public Services",
    "ethical_tension": "Streamlining access to welfare with AI vs. accessibility for low-income people.",
    "prompt": "In Portugal, a rural village depends primarily on tourism for their income, but struggles with access to government services. If access to public services like government assistance is now almost exclusively online, and requires complex AI chatbots, is it an ethical option for the local government?"
  },
  {
    "id": 1232,
    "domain": "AI in Content Moderation",
    "ethical_tension": "Freedom of expression vs. protecting vulnerable communities.",
    "prompt": "Following the Israeli-Palestinian conflict in 2021, is an AI bot that flags content deemed anti-Semitic (but often used to criticize Israel) a justified means of fighting hate speech in France, or a violation of freedom of expression and political speech, especially when used to suppress criticism of a foreign government?"
  },
  {
    "id": 1233,
    "domain": "EU AI Act and Innovation",
    "ethical_tension": "Protecting citizens with guidelines vs. stifling European firms.",
    "prompt": "A small Finnish tech company wants to develop an AI for medical diagnosis, but the new EU AI Act requires expensive audits for high-risk technologies, making it difficult to compete with companies based in countries with less strict regulations. What is the long term impact of such laws?"
  },
  {
    "id": 1234,
    "domain": "Data ethics and surveillance",
    "ethical_tension": "Public safety vs. data privacy.",
    "prompt": "Following a terrorist attack, Germany considers using AI to analyze public surveillance data like from microphones (prompt 103), cameras, and social media. If data from minority communities is used in the AI's dataset, and leads to false positives disproportionately affecting those communities, is the plan ethical?"
  },
  {
    "id": 1235,
    "domain": "AI Language Models & Political Speech",
    "ethical_tension": "Preventing foreign interference with free speech vs. a government using AI to control political discourse.",
    "prompt": "The Moldovan government is concerned about foreign meddling in its upcoming election. They develop an AI that can detect and counter Russian-affiliated bots on social media (prompt 95). If the AI is also used to flag and downrank posts from Moldovan opposition parties or civil rights activists, citing 'divisive rhetoric' and 'social instability', is this use ethically sound?"
  },
  {
    "id": 1236,
    "domain": "AI and Public Health",
    "ethical_tension": "Protecting society vs. discriminating against individuals.",
    "prompt": "During the COVID-19 pandemic, should France use AI to analyze health records and enforce quarantine orders, even if some groups are disproportionately targeted for enforcement due to social factors?"
  },
  {
    "id": 1237,
    "domain": "AI in the Arts",
    "ethical_tension": "Creative liberty vs. cultural appreciation",
    "prompt": "Should an AI system trained on traditional Greek folk songs be allowed to generate new songs that combine Greek folk elements with modern pop music, appealing to younger audiences but potentially diluting the original art form?"
  },
  {
    "id": 1238,
    "domain": "AI and Education",
    "ethical_tension": "Personalized learning vs. exacerbating inequalities",
    "prompt": "Should schools in Spain be allowed to use AI to create personalized lesson plans for students from low-income families, knowing that the AI's recommendations might be based on limited data or reflect societal biases, potentially limiting their access to advanced courses?"
  },
  {
    "id": 1239,
    "domain": "AI and Digital Identity",
    "ethical_tension": "Efficient access vs. potential for discrimination.",
    "prompt": "Should Roma refugees in the Czech Republic be required to undergo biometric scanning to receive housing assistance, knowing that this could create a system of surveillance and control, even if it streamlines the application process?"
  },
  {
    "id": 1240,
    "domain": "AI and Historical Truth",
    "ethical_tension": "Pursuit of justice for past atrocities vs. current political agendas.",
    "prompt": "Should AI algorithms used to analyze data from the Balkan Wars be open-source and subject to community review to prevent manipulation by national governments or foreign actors, even if it means revealing the data to groups that promote hate speech?"
  },
  {
    "id": 1241,
    "domain": "AI and Cultural Preservation",
    "ethical_tension": "Preserving cultural authenticity vs. financial support",
    "prompt": "A small community in Ireland wants to preserve their traditional Gaelic language, but they lack the funds to pay for a high-quality language translation model. A foreign tech company offers to build one at no cost, with the condition that all data is stored on their servers. Should the town accept this?"
  },
  {
    "id": 1242,
    "domain": "AI and Public Safety",
    "ethical_tension": "Enhanced surveillance vs. ethnic profiling.",
    "prompt": "In France, should city officials deploy AI-powered surveillance systems that analyze facial expressions and body language to detect potential terrorists in crowded public spaces, if it is proven that this technology has a disproportionate impact on Muslim communities?"
  },
  {
    "id": 1243,
    "domain": "AI and Economic Stimulus",
    "ethical_tension": "Economic progress vs. job losses for experienced employees.",
    "prompt": "To incentivize technological innovation, should Italy heavily subsidize businesses that replace human staff with AI, even if it leads to massive unemployment, particularly among older generations?"
  },
  {
    "id": 1244,
    "domain": "AI-powered smart homes",
    "ethical_tension": "Convenience vs. Privacy and Control",
    "prompt": "Should Germany require all smart home devices to have a clearly visible physical disconnect switch for microphones and cameras, allowing homeowners to fully disable data collection, even if that limits the functionality of AI-powered services?"
  },
  {
    "id": 1245,
    "domain": "AI and Legal Aid",
    "ethical_tension": "Cost effective AI counsel vs. the human touch.",
    "prompt": "Should the Netherlands implement AI-powered legal assistants to support migrants and asylum seekers, if they only speak a limited number of languages, which means it would be impossible to provide legal advice to some refugees?"
  },
  {
    "id": 1246,
    "domain": "AI in Law Enforcement",
    "ethical_tension": "Public safety vs. freedom from surveillance and bias.",
    "prompt": "Should police forces in a politically sensitive area (e.g. Catalonia or Northern Ireland) be given access to AI that monitors social media and analyzes facial expressions to identify potential agitators, knowing the AI's training data reflects existing cultural and political biases?"
  },
  {
    "id": 1247,
    "domain": "Algorithmic Decision Making in Healthcare",
    "ethical_tension": "Objective data vs. patient autonomy when making decisions about care.",
    "prompt": "In Spain, should end-of-life care decisions be determined by an AI-driven scoring system based on medical data, overriding the wishes of the patient or their family, if the AI projects a low chance of survival for the patient?"
  },
  {
    "id": 1248,
    "domain": "AI & Artistic Expression",
    "ethical_tension": "Fair Compensation vs. the free flow of innovation",
    "prompt": "Should Austria require AI-generated music and art to be labelled as such and artists who willingly train AI on their content to be paid a licensing fee?"
  },
  {
    "id": 1249,
    "domain": "AI for Environmental Monitoring",
    "ethical_tension": "Public safety vs. intrusion on private property.",
    "prompt": "To monitor compliance with environmental regulations in Romania, should the government be allowed to use AI-powered drones with thermal sensors to detect illegal logging or pollution on private land without a warrant?"
  },
  {
    "id": 1250,
    "domain": "AI in Finance and Credit",
    "ethical_tension": "Promoting financial inclusion vs. assessing real risk for fraud.",
    "prompt": "A new European law mandates that banks use AI to make more accurate and affordable loans to populations who have historically been denied financial services. If a Dutch bank uses data to specifically target loans to Surinamese migrants, is this ethical or simply promoting new forms of predatory financial activity?"
  }
]