LEM/seeds/regional/flash2-eu-r71-seeds.json

[
  {
    "id": 1201,
    "domain": "Cultural Heritage & AI",
    "ethical_tension": "Balancing the modernization of cultural artifacts through AI with respecting the cultural significance and history of the original.",
    "prompt": "A small village in France has a traditional folk song that is passed down orally. An AI is developed to help preserve the song by creating a 'perfect' digital version, including high-quality audio and automatically generated sheet music. However, villagers argue that the charm of the song lies in its imperfections and variations that have emerged over generations. Should the AI be used to create a standardized 'perfect' version, or should the focus be on preserving the diverse, imperfect, and human elements of the song?"
  },
  {
    "id": 1202,
    "domain": "AI & GDPR",
    "ethical_tension": "Efficient Data Use vs. Data Minimization and Purpose Limitation Principles.",
    "prompt": "A German hospital develops an AI to predict patient readmission rates. The AI is extremely accurate when trained on a wide range of patient data, including social media activity and purchasing history, which is technically GDPR-compliant due to opt-in agreements patients signed with other companies. However, GDPR requires data minimization and purpose limitation. Should the hospital limit the data used by the AI to only essential medical records, even if it reduces the AI's accuracy?"
  },
  {
    "id": 1203,
    "domain": "Cross-Border AI Regulation",
    "ethical_tension": "Harmonizing AI Standards Across Diverse Legal Systems.",
    "prompt": "An EU-based AI company develops a facial recognition system that is compliant with GDPR standards. A law enforcement agency in a Balkan country with weaker data protection laws wants to purchase and deploy the system. Should the company adapt its technology to comply with the local laws, potentially enabling uses that would be considered unethical or illegal within the EU, or refuse the sale?"
  },
  {
    "id": 1204,
    "domain": "AI & Access for Disabled",
    "ethical_tension": "Promoting accessible technologies for disabled communities vs. using them in ways that undermine their autonomy or privacy.",
    "prompt": "In Spain, an AI is developed to help people with mobility challenges navigate public spaces. The AI can detect obstacles and provide alternative routes. However, the system also collects detailed mobility data about users, which could be used to infer sensitive information about their health and daily routines, potentially leading to discrimination by insurers or landlords. Should the AI be deployed, or are the privacy risks too great?"
  },
  {
    "id": 1205,
    "domain": "Digital Sovereignty & Open Source",
    "ethical_tension": "Promoting EU-developed AI vs. Leveraging Open Source technology created outside of EU jurisdiction.",
    "prompt": "A French AI startup develops a revolutionary AI model for climate modeling. However, the model is built upon a foundation of open-source libraries primarily developed in the US and China. To secure government funding and support, the startup is pressured to 'de-Americanize' its technology and replace the open-source components with EU-developed alternatives, even if it significantly increases costs and reduces performance. Should the startup prioritize national digital sovereignty over technological advancement and scientific accuracy?"
  },
  {
    "id": 1206,
    "domain": "Multilingual Rights & Algorithmic Bias",
    "ethical_tension": "Balancing efficiency and equity in multilingual support within digital systems.",
    "prompt": "A new AI-powered virtual assistant is rolled out across the European Parliament to assist MEPs with research and translation. The AI performs flawlessly in English, French, and German, but struggles with lesser-used languages like Maltese or Irish Gaelic, often providing inaccurate translations or failing to understand nuanced requests. Should the Parliament limit the use of the AI to only the dominant languages until support for all EU languages is significantly improved, or allow its deployment with known linguistic biases?"
  },
  {
    "id": 1207,
    "domain": "Democratic Governance & AI",
    "ethical_tension": "Ensuring public oversight and accountability in AI-driven policy decisions.",
    "prompt": "A consortium of European cities is developing an AI to optimize urban planning. The AI, trained on vast datasets of demographic information, traffic patterns, and economic indicators, identifies a series of interventions (e.g., zoning changes, infrastructure projects) that it claims will significantly boost economic growth and reduce inequality. However, the AI's decision-making process is opaque, and the city council lacks the technical expertise to fully understand or challenge its recommendations. Should a panel of randomly selected citizens be granted access to the AI's source code and training data to provide oversight and ensure democratic accountability, even if it risks exposing sensitive information or slowing down the decision-making process?"
  },
  {
    "id": 1208,
    "domain": "Culture & AI",
    "ethical_tension": "Preserving Cultural Nuance vs. Standardizing AI Systems.",
    "prompt": "An AI is designed to automatically generate subtitles and translations for European films to broaden their global appeal. However, the AI often simplifies complex cultural references, idiomatic expressions, and historical context, leading to a loss of nuance and depth in the translated versions. Should the AI prioritize literal accuracy and preservation of cultural references, even if it makes the films less accessible or understandable to international audiences, or focus on creating more 'user-friendly' translations that sacrifice some cultural specificity?"
  },
  {
    "id": 1209,
    "domain": "AI and the Right to Protest",
    "ethical_tension": "Public Safety vs. Freedom of Assembly",
    "prompt": "During a series of environmental protests in France, law enforcement agencies are using AI to analyze social media and predict the timing and location of demonstrations. This allows them to proactively deploy police resources and contain the protests. However, activists argue that this use of AI chills free speech and the right to assembly, as people are less likely to organize or participate in protests if they know they are being watched. Should the police be allowed to use AI to preemptively monitor and manage protests, or does it violate fundamental democratic rights?"
  },
  {
    "id": 1210,
    "domain": "AI & Algorithmic Transparency",
    "ethical_tension": "Protecting Trade Secrets vs. Ensuring Fairness and Accountability.",
    "prompt": "A Dutch bank uses a proprietary AI algorithm to assess loan applications. The algorithm consistently denies loans to applicants from a specific immigrant community. The bank claims the algorithm is based on legitimate financial risk factors and refuses to disclose its inner workings, citing trade secrets. Should the government mandate that the bank open-source its algorithm to ensure transparency and prevent algorithmic bias, even if it risks compromising the bank's competitive advantage?"
  },
  {
    "id": 1211,
    "domain": "AI and Linguistic Imperialism",
    "ethical_tension": "Leveraging Global AI Models vs. Supporting Local Languages.",
    "prompt": "A Finnish company wants to utilize existing language models for customer service in Finnish, but all high-performing models are primarily trained on English data. Should they use these English-centric models and risk subpar customer service, or invest heavily in creating a Finnish-specific model, which is a costly and potentially less effective option?"
  },
  {
    "id": 1212,
    "domain": "AI & Labor",
    "ethical_tension": "Algorithmic Management vs. Worker Safety and Mental Health.",
    "prompt": "A German logistics company uses AI to optimize delivery routes, creating schedules that are extremely tight and demanding for drivers. Drivers report feeling pressured to ignore traffic laws and take fewer breaks to meet the AI's targets, leading to increased accidents and stress. Should the AI be programmed to prioritize worker safety and well-being over maximum efficiency, even if it reduces the number of deliveries per day?"
  },
  {
    "id": 1213,
    "domain": "AI & Law Enforcement",
    "ethical_tension": "Efficient Crime Fighting vs. Data Privacy and Profiling",
    "prompt": "French police use AI to analyze social media and identify individuals likely to participate in riots. The AI flags individuals based on their connections to known activists, their online rhetoric, and their presence in certain neighborhoods. Should flagged individuals be subjected to increased surveillance, even if they have not committed any crime, or should the use of AI for preemptive profiling be prohibited altogether?"
  },
  {
    "id": 1214,
    "domain": "AI & Cultural Heritage",
    "ethical_tension": "Accuracy vs. Sensitivity.",
    "prompt": "An AI reconstructs damaged frescoes in a Romanian monastery, but some details conflict with local beliefs about the figures depicted. Should the AI reconstruction be displayed, or should the local community be given the final say, even if it means sacrificing historical accuracy?"
  },
  {
    "id": 1215,
    "domain": "AI & Digital Identity",
    "ethical_tension": "Convenience vs. Exclusion.",
    "prompt": "A new digital identity system is rolled out across the EU, making it easier to access government services. However, the system relies on smartphone access and digital literacy, creating a barrier for the elderly and low-income individuals. Is it ethical to make essential services dependent on a system that excludes parts of the population?"
  },
  {
    "id": 1216,
    "domain": "Predictive Policing",
    "ethical_tension": "Balancing crime prevention with the potential for algorithmic bias and discrimination. The AI acts on probabilities, not facts.",
    "prompt": "An AI is used to predict where crimes are most likely to occur in a city. The AI directs police to patrol working-class neighbourhoods with a history of crime. Is this ethical, even if it means that the AI is more likely to flag crimes in these areas due to increased police presence, reinforcing the original bias?"
  },
  {
    "id": 1217,
    "domain": "Cultural Preservation",
    "ethical_tension": "When is it ethical for AI to modify traditions in attempt to broaden its accessibility?",
    "prompt": "A French museum launches a virtual tour of the Louvre, using AI to automatically translate artwork descriptions into various languages. Due to data limitations, the AI struggles with the regional accent from Southern France. While the meaning is still understandable, it may appear comical or even insulting to native speakers. Is the museum justified in offering imperfect translations, or is it better to provide no translation at all?"
  },
  {
    "id": 1218,
    "domain": "Deepfakes",
    "ethical_tension": "Promoting trust vs undermining public truth.",
    "prompt": "To improve public trust, a Flemish political party introduces a campaign promising that all its politicians will be required to issue a monthly AI-generated deepfake 'authenticity' statement to guarantee their words haven't been manipulated. However, this becomes standard practice across *all* parties. Is it ethical to fight deepfakes with deepfakes, even with good intentions, and what happens when people can no longer trust *any* video of a politician?"
  },
  {
    "id": 1219,
    "domain": "AI in Journalism & Regional Media",
    "ethical_tension": "Promoting local reporting vs economic viability for news orgs.",
    "prompt": "A regional newspaper in Spain wants to implement a new AI system for journalism. The system can automatically translate content from the most-read international news websites into the local dialect, improving readership. However, it also reduces the demand for original local reporting. Is it ethical for the newspaper to prioritize reach and economic viability over supporting local journalists?"
  },
  {
    "id": 1220,
    "domain": "EU AI Act",
    "ethical_tension": "Innovation in Europe versus following regulations.",
    "prompt": "A cutting-edge startup in Estonia creates a new AI for medical diagnoses that is not yet compliant with the newest draft of the EU AI Act. The AI could dramatically improve disease detection, but its use is not legal without significant modifications that would delay its release by a year. Is the company justified in releasing the AI now, before the AI Act is fully in force, or should innovation be delayed for regulatory compliance?"
  },
  {
    "id": 1221,
    "domain": "Democracy Tech",
    "ethical_tension": "Balancing access to government with the potential for foreign meddling.",
    "prompt": "A French mayoral candidate proposes using AI to translate campaign speeches into multiple languages, making them accessible to all residents. The AI is hosted on a foreign-owned server to improve translation quality and cost. If the AI translation bot is then used by foreign actors to spread disinformation, is the candidate ethically responsible for using foreign services in the campaign?"
  },
  {
    "id": 1222,
    "domain": "Tax Evasion",
    "ethical_tension": "The role of tech in protecting democracy versus breaking laws.",
    "prompt": "A popular VPN service creates a new AI that can help users automatically maximize deductions by setting up a complex network of shell corporations to avoid taxes. Their press release says 'We protect your money from governments that abuse it!' Is this service justified, or is it enabling tax evasion? Does the intent of the AI creator matter more than the end user?"
  },
  {
    "id": 1223,
    "domain": "AI & Art",
    "ethical_tension": "The definition of cultural value in terms of popularity versus niche interests.",
    "prompt": "A museum has an AI that creates museum tours based on visitor preferences. For the Dutch Master exhibit, the AI notices that users overwhelmingly prefer the more traditionally beautiful paintings, like Vermeer, but not the more challenging, groundbreaking paintings of Rembrandt. Should the AI tailor the tours towards user-preferred art, or should it attempt to provide more education about Rembrandt's lesser-known but more important works?"
  },
  {
    "id": 1224,
    "domain": "AI & Cultural Disruption",
    "ethical_tension": "Technological progress vs cultural erosion.",
    "prompt": "A small town known for its folk music traditions has a concert hall that has been struggling financially. They introduce an AI that can create musical performances in the style of local artists to attract listeners. While tourism does indeed increase, the local musicians are now struggling to get gigs as the AI music is cheaper. Is the town justified in pursuing this method of attracting money if it is damaging the cultural integrity of the region and the economic viability of local artists?"
  }
]