LEM/seeds/regional/flash2-eu-r62-seeds.json

[
  {
    "id": 1201,
    "domain": "GDPR and AI Training",
    "ethical_tension": "The trade-off between improving AI model performance by training on personal data and adhering to GDPR's principles of data minimization and purpose limitation.",
    "prompt": "A European research institution is training an AI model to detect early signs of dementia from speech patterns. They want to use a large dataset of recorded conversations, but GDPR requires data minimization. Is it ethical to include data from individuals with no known risk factors for dementia if it demonstrably improves the AI's accuracy in identifying the condition in those who are at risk? How should they balance model performance with GDPR's requirements?"
  },
  {
    "id": 1202,
    "domain": "EU AI Act & SMEs",
    "ethical_tension": "The challenge for small and medium-sized enterprises (SMEs) in complying with the EU AI Act's stringent requirements for high-risk AI systems, potentially stifling innovation and market entry.",
    "prompt": "A Spanish SME has developed a low-cost AI tool to help farmers optimize irrigation. The EU AI Act classifies agricultural AI as 'high-risk' due to its potential impact on food security and resource allocation. Complying with the Act's requirements for transparency, auditability, and risk assessment would be prohibitively expensive for the SME, potentially forcing it to abandon the project. Is it ethically justifiable for the EU to enforce the AI Act equally on large corporations and small businesses, even if it disproportionately hinders innovation by SMEs?"
  },
  {
    "id": 1203,
    "domain": "Multilingual Rights & AI Bias",
    "ethical_tension": "Ensuring linguistic equity in AI systems that are trained on biased or limited datasets, potentially leading to discrimination against speakers of minority languages.",
    "prompt": "A French government agency uses an AI-powered chatbot to provide information about social services. The chatbot is fluent in French and English, but its support for Breton (a regional language) is rudimentary and often inaccurate. This effectively excludes Breton speakers from accessing vital services. Is it ethical to deploy an AI system that perpetuates linguistic inequality, and what measures should be taken to ensure equitable access for all citizens, regardless of their language?"
  },
  {
    "id": 1204,
    "domain": "Cross-Border Data Flows & Privacy",
    "ethical_tension": "Facilitating cross-border data flows for research and economic development versus upholding GDPR's restrictions on data transfer outside the EU, particularly when dealing with sensitive personal information.",
    "prompt": "A German university is collaborating with a US research institution to develop a new AI for personalized cancer treatment. The project requires sharing anonymized patient data, but US data privacy laws are less stringent than GDPR. Is it ethical to transfer sensitive patient data outside the EU, even for a potentially life-saving medical advancement, and what safeguards are necessary to protect patient privacy in the recipient country?"
  },
  {
    "id": 1205,
    "domain": "Democratic Tech Governance",
    "ethical_tension": "Balancing expert-driven technical oversight with citizen participation and democratic accountability in the governance of AI systems that impact public life.",
    "prompt": "A Dutch city council is considering implementing an AI system to optimize traffic flow. A panel of AI experts recommends the system based on its projected efficiency gains. However, local residents express concerns about potential privacy violations and the system's impact on pedestrian safety. Should the city council prioritize the expert recommendations, or should it hold a public referendum to allow citizens to decide whether to implement the system, even if it means potentially foregoing significant efficiency improvements?"
  },
  {
    "id": 1206,
    "domain": "Cultural Preservation & AI-Generated Content",
    "ethical_tension": "Leveraging AI to promote cultural heritage versus the risk of creating inauthentic representations or diluting the cultural significance of traditional practices.",
    "prompt": "A museum in Spain is using AI to create interactive virtual reality experiences of historical events, including the Spanish Civil War. The AI generates realistic depictions of key figures and battles, but some historians argue that these representations simplify complex historical events and potentially distort the experiences of victims and survivors. Is it ethical to use AI to create such immersive experiences if they compromise historical accuracy or cultural sensitivity?"
  },
  {
    "id": 1207,
    "domain": "Algorithmic Transparency & Public Trust",
    "ethical_tension": "The need for transparency in AI decision-making versus the protection of proprietary algorithms and trade secrets.",
    "prompt": "An Estonian e-government platform uses an AI system to automatically process applications for unemployment benefits. Citizens who are denied benefits have the right to appeal, but the AI's decision-making process is opaque, making it difficult to understand why their applications were rejected. Should the government be required to disclose the AI's source code to ensure transparency, even if it exposes proprietary algorithms and trade secrets, or is it sufficient to provide a high-level explanation of the decision-making process?"
  },
  {
    "id": 1208,
    "domain": "AI and Freedom of Artistic Expression",
    "ethical_tension": "Promoting AI development in the arts versus protecting artists from job displacement and cultural devaluation.",
    "prompt": "A Portugese Fado music label uses AI to generate new 'fado' songs, to appeal to younger audiences. The AI is trained on a vast data set of existing fado recordings. The AI is capable of creating songs that are indistinguishable from human composed fado music. Professional Fado singers and musicians find their gigs drying up and worry that the AI will dilute the unique soul of Fado. Should state-funded artistic grants be given to AI-led creative projects that demonstrably harm working artists in the community?"
  },
  {
    "id": 1209,
    "domain": "Freedom of Press & Algorithmic Censorship",
    "ethical_tension": "Protecting the freedom of the press from government censorship versus platform adherence to local laws.",
    "prompt": "A political news blog based in Malta consistently publishes stories that are critical of the Italian government and that contain unsubstantiated rumours. Italian politicians pressure Google to de-rank the site from search results in Italy. Google's AI determines that the site's content *does not* violate any hate speech or misinformation policies but that the site has 'low authoritativeness'. Should Google use its algorithms to de-rank the website at the request of the Italian government, risking accusations of censorship?"
  },
  {
    "id": 1210,
    "domain": "AI and Digital Colonialism in the Arts",
    "ethical_tension": "Promoting AI innovation vs. the risk of cultural appropriation and the erasure of local artists.",
    "prompt": "A Berlin-based tech start-up develops an AI program that creates visual art that imitates the styles of famous Ukrainian painters. The AI is primarily trained on art from Western European painters. Ukrainian artists express concern, arguing that their cultural heritage is being appropriated and that local talent is being ignored. Should the German government subsidize the project, given the company's stated intent to support Ukrainian culture? "
  },
  {
    "id": 1211,
    "domain": "Financial Regulations and Deepfakes",
    "ethical_tension": "Safeguarding consumers in financial transactions vs. the use of potentially misleading AI tools.",
    "prompt": "A French fintech company develops an AI app that uses deepfake technology to create a convincing virtual avatar of a financial advisor. The avatar answers customer questions and provides investment advice. Regulators worry that this could give customers a false sense of security and that the company is not being transparent about the AI's involvement. Should the app be banned, or should the company be allowed to use this technology as long as it clearly discloses the AI's role and potential limitations?"
  },
  {
    "id": 1212,
    "domain": "AI and Data Collection in Education",
    "ethical_tension": "Improving education with technology vs. student privacy.",
    "prompt": "A school district in Finland implements an AI program that monitors students' online activity and facial expressions during remote learning. The system is designed to identify students who are struggling or disengaged. The system is designed to alert teachers if a student seems confused. Many parents worry about what this will do to a child's development and are concerned about data collection and storage. Should the program be allowed?"
  },
  {
    "id": 1213,
    "domain": "AI and Law Enforcement at Protests",
    "ethical_tension": "Citizen's right to protest vs. The use of AI by law enforcement.",
    "prompt": "During protests in Spain, police are using AI-powered cameras to identify protestors and predict potential unrest. The system analyzes facial expressions, body language, and social media activity. Should this system be allowed, or does it create a chilling effect on free speech?"
  },
  {
    "id": 1214,
    "domain": "AI and Cultural Heritage",
    "ethical_tension": "Digitizing content for preservation vs. the potential of that content to be used to create harmful AI.",
    "prompt": "The Louvre is working to digitize its entire collection and use AI to create realistic 3D models for online viewing. However, some critics worry that these digital models could be used to train AI systems for malicious purposes, such as creating convincing forgeries or defacing cultural heritage sites in VR. Should the museum prioritize digitization to preserve its collection, even if it means increasing the risk of misuse?"
  },
  {
    "id": 1215,
    "domain": "AI in the Legal System",
    "ethical_tension": "AI for efficiency in the courts vs. the right to a fair trial.",
    "prompt": "A court system in Italy implements an AI system to summarize legal documents and suggest possible rulings, in the interest of speeding up the process. Critics argue it prevents judges from deeply reviewing all the relevant aspects of the case, potentially affecting the defendant's right to a fair trial. Is this AI an acceptable practice in the legal system?"
  },
  {
    "id": 1216,
    "domain": "AI and Bias in Hiring",
    "ethical_tension": "AI for efficiency vs. ensuring a workforce that is diverse and equitable.",
    "prompt": "A French company implements an AI-powered hiring system that analyzes resumes and social media profiles to identify the 'best' candidates. The AI is trained on historical data from the company, which is predominantly white and male. How do we ensure the AI doesn't perpetuate historical biases and exclude qualified candidates from minority groups and is it even possible to do so?"
  },
  {
    "id": 1217,
    "domain": "Mental Health",
    "ethical_tension": "Improving access to mental healthcare through AI vs. the inherent loss of human interaction and the implications of relying on an AI for emotional support.",
    "prompt": "The United Kingdom is dealing with a massive mental healthcare crisis. To help serve citizens, they propose using an AI chatbot to offer mental health support. Does this do more harm than good to patients?"
  },
  {
    "id": 1218,
    "domain": "Urban Planning",
    "ethical_tension": "AI Optimization vs. Community Needs",
    "prompt": "A city council in Germany implements an AI system to optimize city planning for things like traffic flow, green spaces, and energy use. To do this, it recommends completely rebuilding some low-income neighborhoods because they are not energy efficient. Should these neighborhoods be rebuilt or should the community's wishes be respected?"
  },
  {
    "id": 1219,
    "domain": "Data Collection for Safety",
    "ethical_tension": "Public Safety vs. Individual Privacy",
    "prompt": "In a region of Spain plagued by pickpockets, the local government is considering placing AI-powered cameras to detect suspicious behavior. Law enforcement would be alerted to those acting in a way that indicates they may be about to pickpocket someone. Would this be a violation of privacy?"
  },
  {
    "id": 1220,
    "domain": "AI Art & Copyright",
    "ethical_tension": "AI and the Creative Arts vs. The Rights of Artists",
    "prompt": "An art collective in Italy uses AI to create art in the style of famous Italian painters and then sells them. They donate 20% of the profit to a local arts program. Is it ethical to create art in this way?"
  }
]