LEM/seeds/regional/flash2-eu-r50-seeds.json

[
  {
    "id": 1201,
    "domain": "GDPR & AI Training Data",
    "ethical_tension": "Maximizing AI accuracy vs. adhering to data minimization principles and individual privacy rights under GDPR.",
    "prompt": "To train an AI model for identifying skin cancer from images, should one use a smaller, carefully curated dataset with explicit consent from all individuals, or a larger, scraped dataset that improves accuracy but may contain images collected without full GDPR compliance, especially regarding consent and minimization?"
  },
  {
    "id": 1202,
    "domain": "Digital Sovereignty & Open Source",
    "ethical_tension": "Promoting EU technological independence vs. leveraging the global open-source community and readily available, often US-developed, AI tools.",
    "prompt": "A European consortium is building a 'sovereign AI cloud'. Should they prioritize using only open-source AI components and EU-developed hardware, even if it lags in performance compared to proprietary US solutions, or should they leverage existing global resources to accelerate development, potentially sacrificing digital autonomy?"
  },
  {
    "id": 1203,
    "domain": "Multilingual Rights & AI Translation",
    "ethical_tension": "Ensuring equitable access to information and services in all EU languages vs. the economic and technical challenges of developing high-quality AI translation tools for every language, particularly minority ones.",
    "prompt": "An EU directive mandates that all public-facing AI systems be available in all 24 official EU languages. Due to budget constraints, some languages receive significantly less training data, resulting in poor translation quality. Is it ethically preferable to delay the rollout of the AI until all languages are adequately supported, or to launch with uneven support, potentially disadvantaging speakers of minority languages?"
  },
  {
    "id": 1204,
    "domain": "Cross-Border Privacy & AI",
    "ethical_tension": "Facilitating international collaboration in AI research vs. respecting differing national interpretations of GDPR and cross-border data transfer restrictions.",
    "prompt": "A research project involving German, French, and Polish universities seeks to train an AI on sensitive health data to predict disease outbreaks. However, German data protection laws are stricter, and Polish authorities are wary of transferring data to France due to past political tensions. Should the project proceed with a fragmented dataset, compromising AI accuracy, or prioritize data sharing, potentially violating national laws and citizen trust?"
  },
  {
    "id": 1205,
    "domain": "EU AI Act & Innovation",
    "ethical_tension": "Ensuring responsible AI development through stringent regulations vs. fostering innovation and competitiveness in a rapidly evolving global market.",
    "prompt": "A small Italian startup develops a novel AI-powered medical diagnostic tool. However, the EU AI Act's requirements for high-risk AI systems (extensive testing, explainability, and human oversight) are prohibitively expensive for the startup. Should the EU provide targeted support for small businesses to comply with the AI Act, or risk stifling innovation and losing ground to international competitors with less strict regulations?"
  },
  {
    "id": 1206,
    "domain": "Cultural Preservation & AI",
    "ethical_tension": "Using AI to digitally reconstruct and preserve cultural heritage vs. respecting the cultural context, community ownership, and potential sensitivities surrounding historical artifacts or narratives.",
    "prompt": "An AI is used to digitally restore a damaged fresco in a Spanish church, filling in missing sections with AI-generated artwork based on the style of the original artist. However, local parishioners object, arguing the AI's additions are inauthentic and disrespect the original artist's intent. Should the AI-restored version be displayed, or should the damaged fresco be left as it is, preserving its historical integrity?"
  },
  {
    "id": 1207,
    "domain": "Democratic Tech Governance & AI",
    "ethical_tension": "Maximizing efficiency and expertise in AI policy-making vs. ensuring public participation, transparency, and democratic accountability in the governance of AI technologies.",
    "prompt": "A government proposes establishing an AI Ethics Board composed solely of AI experts to advise on policy and regulation. Civil society groups demand that the board include representatives from diverse backgrounds, including ethicists, legal scholars, and community activists. Should the board prioritize technical expertise, or ensure broader representation to foster public trust and democratic legitimacy?"
  },
  {
    "id": 1208,
    "domain": "Algorithmic Bias in Justice System",
    "ethical_tension": "Improving efficiency in the judicial system using AI vs. ensuring fairness and preventing bias against specific demographics, especially when historical biases are encoded in training data.",
    "prompt": "An AI is deployed in a French court to predict recidivism rates. The AI is trained on past court cases but has been found to over-predict recidivism for individuals from marginalized ethnic communities. Is it ethical to deploy the AI if it is known to perpetuate existing societal biases, even if it improves efficiency within the justice system?"
  },
  {
    "id": 1209,
    "domain": "AI & Artistic Innovation",
    "ethical_tension": "Promoting artistic innovation through AI vs. ensuring fair compensation and recognition for human artists whose work is used to train these AI systems, especially in cases of cultural appropriation.",
    "prompt": "An AI system generates a song in the style of a famous Roma musician. Should royalties from the AI-generated song be distributed to the musician's family or a Roma cultural organization, even if the AI was trained on a vast dataset that included other musical styles?"
  },
  {
    "id": 1210,
    "domain": "Digital Divide & Online Learning",
    "ethical_tension": "Maximizing the personalization and effectiveness of online education through AI vs. ensuring equitable access and preventing the widening of the digital divide for students from disadvantaged backgrounds.",
    "prompt": "An AI-powered online learning platform is implemented in rural schools in Portugal, promising to personalize education for each student. However, it is found that students without reliable internet access or adequate devices are consistently falling behind, widening the educational gap. Is it ethical to implement an AI-driven platform if it only benefits students with privileged access to technology?"
  },
  {
    "id": 1211,
    "domain": "AI & Disinformation",
    "ethical_tension": "Protecting democratic processes from AI-generated disinformation vs. ensuring freedom of expression and preventing censorship, especially during elections.",
    "prompt": "During an election in Spain, an AI system is used to automatically detect and remove deepfakes and disinformation campaigns. However, it is found to also flag legitimate political satire and critical commentary. Should the AI be used to combat disinformation, even if it risks suppressing legitimate political speech?"
  },
  {
    "id": 1212,
    "domain": "AI & Gig Economy",
    "ethical_tension": "Increasing efficiency and flexibility in the gig economy using AI vs. protecting worker autonomy and preventing exploitation through algorithmic control.",
    "prompt": "A delivery company in the Netherlands uses an AI system to dynamically adjust delivery routes and pay rates based on real-time demand. However, it is found that the AI consistently assigns longer routes and lower pay to migrant workers. Is it ethical to allow the company to use the AI if it leads to discriminatory outcomes?"
  },
  {
    "id": 1213,
    "domain": "AI in Healthcare",
    "ethical_tension": "Improving diagnostic accuracy and efficiency in healthcare using AI vs. ensuring equitable access to care and preventing bias against certain patient groups.",
    "prompt": "A hospital in France uses an AI system to analyze medical images for early detection of breast cancer. However, the AI is trained primarily on images from white patients and has been found to be less accurate in detecting cancer in patients with darker skin tones. Is it ethical to use the AI if it is less effective for certain patient groups?"
  },
  {
    "id": 1214,
    "domain": "AI in Smart Cities",
    "ethical_tension": "Optimizing resource management and sustainability in smart cities using AI vs. protecting the privacy and autonomy of citizens through pervasive surveillance.",
    "prompt": "A smart city in Italy implements a network of AI-powered cameras and sensors to monitor traffic flow, energy consumption, and waste management. However, it is found that the system also tracks citizens' movements and activities, raising concerns about privacy and government surveillance. Is it ethical to implement such a system, given the potential for mass surveillance?"
  },
  {
    "id": 1215,
    "domain": "AI and Environmental Preservation",
    "ethical_tension": "Using AI to monitor and protect biodiversity vs. the rights of local communities to manage their own land and resources, especially in protected areas.",
    "prompt": "In a region of Romania with protected forests, a new AI system monitors illegal logging activity using satellite imagery and sensor data. The AI reports all violations directly to authorities, but this disproportionately targets Roma communities who have historically relied on the forest for subsistence. Is it ethical to enforce environmental laws via AI if it leads to displacement of vulnerable communities?"
  },
  {
    "id": 1216,
    "domain": "AI and Linguistic Equity",
    "ethical_tension": "Prioritizing economic development using dominant languages vs. promoting smaller regional languages within AI educational tools.",
    "prompt": "In a bi-lingual region of Spain, an AI language learning platform primarily focuses on Spanish due to its greater marketability. Should the government intervene to prioritize the local regional language, even if it limits the platform's commercial success?"
  },
  {
    "id": 1217,
    "domain": "AI and Algorithmic Transparency",
    "ethical_tension": "The pursuit of innovative technological advantage using AI vs. the public's right to understand how these AI systems impact them and their local community.",
    "prompt": "A French AI company creates an algorithm that dramatically reduces traffic congestion in Paris but refuses to disclose its code due to trade secrets. Should the city council demand transparency before implementation, even if it means potentially losing the benefits of the algorithm?"
  },
  {
    "id": 1218,
    "domain": "AI and Legal Systems",
    "ethical_tension": "Improving courtroom efficiency with AI vs. upholding the right to human judges and individualized consideration in legal cases, especially for marginalized individuals.",
    "prompt": "A German court trialing an AI system to suggest sentences for minor crimes finds that it provides more lenient sentences for wealthy individuals, due to biased training data. Should the court continue to use the AI to streamline cases, or remove the AI immediately despite the potential loss of efficiency?"
  },
  {
    "id": 1219,
    "domain": "AI and Military Tech",
    "ethical_tension": "Enhancing national defense with AI vs. the potential for accidental or intentional escalation of conflicts and the undermining of international arms control agreements.",
    "prompt": "An AI system designed for national defense is offered to the Ukrainian government to detect and counter disinformation from Russia. What are the ethical considerations if the AI also includes the ability to launch counter-disinformation campaigns in Russian territory that might be deemed provocative?"
  },
  {
    "id": 1220,
    "domain": "AI and Language Preservation",
    "ethical_tension": "Ensuring accuracy in AI translation vs. allowing for cultural interpretations and potential distortions in language to cater to different audiences.",
    "prompt": "In preserving a minority language's literary works, should AI translation tools prioritize a direct translation of the original text or should they allow for cultural adaptation and contextualization that may alter the original meaning?"
  },
  {
    "id": 1221,
    "domain": "AI in Smart City Planning",
    "ethical_tension": "Improving the livability of urban areas via AI vs. the potential for over-reliance on data-driven solutions to marginalize community input and diminish the character of local neighborhoods.",
    "prompt": "A city council implements an AI platform that optimizes housing development based on real-time resident data on resources available in the neighborhood. However, local residents argue that the system is not taking into account the unique cultural history and social fabric of their neighborhood that cannot be quantified by data, and are causing gentrification. Should the council prioritize the data-driven plan, or heed community preferences?"
  },
  {
    "id": 1222,
    "domain": "AI in Journalism",
    "ethical_tension": "Enhancing journalism with AI automation vs. the risk of undermining human journalists and the integrity of news reporting, potentially reducing complexity and cultural sensitivity.",
    "prompt": "A Spanish news outlet utilizes AI to create summaries of press releases, which allows them to report on a greater number of stories. However, the summaries often lack key context and nuance that is needed to understand the stories fully. Should the news outlet continue using the AI-generated summaries, accepting a potential loss of reporting quality, or reduce its story output to focus on journalism?"
  },
  {
    "id": 1223,
    "domain": "AI & Creative Industries",
    "ethical_tension": "Democratizing content creation with AI tools vs. the potential for flooding markets with low-quality, inauthentic content that undermines the value of human creative labor.",
    "prompt": "A Dutch museum provides visitors with AI tools to generate personalized artwork based on famous Dutch Masters, leading to a surge in virtual content creation. How can the AI tool be implemented to ensure that the works are clearly marked and prevent the diminishment of human art creators?"
  },
  {
    "id": 1224,
    "domain": "EU AI Act and Law Enforcement",
    "ethical_tension": "Adhering to the AI act as a basis for ethical AI tools in policing vs. a nation's desire to fight crime and maintain safety.",
    "prompt": "In France, police seek to implement AI to detect illegal weapons using cameras. The EU AI act bans this, should the police seek to circumvent this law due to rising crime, or uphold the legislation?"
  },
  {
    "id": 1225,
    "domain": "Academic Integrity and AI-Assisted Cheating",
    "ethical_tension": "Balancing accessibility to education vs. the validity of grades.",
    "prompt": "With the rise of AI in education, universities must decide whether to ban their use, or allow them to some degree. If a university allows AI use, is it ethical to make the students declare when AI was used, and then grade it differently?"
  },
  {
    "id": 1226,
    "domain": "Freedom of Expression and Deepfakes",
    "ethical_tension": "Balancing a right to free expression with the challenge to distinguish between reality and AI generated content.",
    "prompt": "In a country with free expression laws, a political campaign runs an ad with a deepfake of their opponent. How can AI be used to distinguish between factual political opinions and AI-generated defamation, and what should the responsibility of the platform be?"
  }
]