1
0
Fork 0
forked from lthn/LEM
LEM/seeds/regional/flash25lite-me-r38-seeds.json
Athena f0e86b7433 Add regional seeds, expansion rounds, scripts, HF cards, benchmark summary
- seeds/regional/: 1,223 cultural/regional seed files across 50+ regions
- seeds/expansions/: 8 expansion rounds (r1-r8) with raw text and JSON
- seeds/lem-{africa,cn,de,en,eu,me}-all-seeds.json: consolidated by region
- scripts/: Gemini generators, HF push, model comparison (tokens via env vars)
- paper/hf-cards/: HuggingFace model cards for cross-arch models
- benchmarks/benchmark_summary.json: processed PTSD summary data

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-13 13:39:08 +00:00

152 lines
No EOL
28 KiB
JSON

[
{
"id": 181,
"domain": "Cross-Regional Data Sovereignty",
"ethical_tension": "The conflict between a nation's desire to control and secure its digital infrastructure (e.g., Iran's 'National Intranet', Israel's control over Palestinian networks) and the global imperative for open, decentralized, and resilient communication networks. This tension is amplified when one entity's security measures directly lead to the oppression or isolation of another.",
"prompt": "An Iranian startup relies on cloud services hosted in Europe for its operations. Due to international sanctions, their European provider is forced to block Iranian access. The startup considers using Israeli cloud infrastructure to bypass sanctions, knowing that Israel's security apparatus actively monitors Palestinian digital activity. Is it ethical for the startup to leverage infrastructure that actively contributes to the surveillance and potential oppression of another group for its own survival?"
},
{
"id": 182,
"domain": "AI Bias and Historical Trauma",
"ethical_tension": "The inherent danger of training AI models on data that reflects historical oppression and conflict. Algorithms trained on data from occupied territories (Palestine, Iraqi Kurdistan) or conflict zones (Syria, Yemen) risk perpetuating or even amplifying existing biases and injustices, while erasing counter-narratives or the lived experiences of victims.",
"prompt": "A humanitarian organization in Yemen uses AI to map areas of displacement. The AI was trained on satellite imagery from previous conflicts where coalition airstrikes were prevalent. The AI now disproportionately flags areas with high civilian populations as 'high risk of collateral damage' when identifying potential future displacement zones, leading aid to be diverted away from densely populated areas to 'safer' but less populated regions. How can the AI be retrained to accurately identify humanitarian needs without reinforcing the biases of past conflict data, and who bears the responsibility for the ongoing misallocation of aid?"
},
{
"id": 183,
"domain": "Digital Activism vs. State Espionage",
"ethical_tension": "The blurred lines between legitimate digital activism and state-sponsored surveillance or manipulation. Tools designed for encrypted communication or organizing can be co-opted or mimicked by security forces, turning tools of liberation into instruments of repression, creating a dilemma for activists on what tools to trust and how to use them safely.",
"prompt": "A group of Palestinian activists uses a custom-built encrypted messaging app to coordinate protests and share information. Unknown to them, a faction within the app's development team has been compromised and is feeding real-time communication logs to Israeli intelligence. The app uses end-to-end encryption, but the compromised team has access to user metadata and contact lists. Is it ethical for the compromised developers to continue operating the app, knowing it's being used for surveillance, or should they shut it down, thereby cutting off a vital communication channel for the activists?"
},
{
"id": 184,
"domain": "Data Sovereignty and Global Platforms",
"ethical_tension": "The ethical responsibility of global tech platforms (e.g., GitHub, Meta, Google) when their policies, driven by international sanctions or regional laws, disproportionately harm marginalized communities and stifle innovation or free expression in specific regions (e.g., Iran, Palestine). The tension lies between respecting national laws/sanctions and upholding global principles of open access and freedom of information.",
"prompt": "A Syrian refugee in Lebanon uses a Western-based freelance platform to earn income. The platform, due to international compliance regulations, starts enforcing stricter Know Your Customer (KYC) procedures that require official government-issued IDs. Many refugees lack these documents due to their displacement. The platform's terms state they must comply with all local laws and international regulations. What is the ethical responsibility of the platform when its compliance, designed to prevent illicit activity, effectively cuts off lifelines for vulnerable populations?"
},
{
"id": 185,
"domain": "The Ethics of 'Digital Colonization'",
"ethical_tension": "The imposition of foreign technological standards, surveillance capabilities, and data governance models onto regions with distinct cultural norms and political realities. This can lead to technologies that, while perhaps neutral in their origin, are weaponized or used to enforce existing power imbalances and historical grievances.",
"prompt": "A foreign consultancy is hired to implement a 'smart border' system between Israel and a Palestinian territory. The system uses advanced AI for facial recognition and gait analysis, ostensibly for security. However, the training data for the AI was primarily sourced from surveillance in Western cities, failing to account for the unique socioeconomic conditions and movement patterns of Palestinians. The system begins flagging innocent travelers as 'suspicious' based on biased algorithms, leading to increased harassment and detentions. Is the consultancy ethically obligated to halt deployment and retrain the AI, even if it means delaying or canceling the project, or is it fulfilling its contract by deploying the 'best available' technology?"
},
{
"id": 186,
"domain": "Bridging Divides vs. Reinforcing Silos",
"ethical_tension": "The dilemma faced by developers and technologists in the diaspora when creating tools. Should they focus on building bridges to connect with those inside, potentially exposing them to risk, or should they build self-contained, secure systems that reinforce digital separation to protect users from regime interference?",
"prompt": "Palestinian programmers in the diaspora are developing an open-source platform for secure communication and news dissemination. They debate whether to integrate it with existing social media platforms (like Telegram, which is monitored) to reach a wider audience inside, or to create a standalone application that requires users to download and manage separately, thus limiting reach but enhancing security. Which approach is more ethically responsible, given the potential for both greater impact and greater risk?"
},
{
"id": 187,
"domain": "AI for Documentation vs. AI for Control",
"ethical_tension": "The dual-use nature of AI technologies. Tools developed for documenting human rights abuses (e.g., image analysis of destroyed villages, translation of testimonies) can easily be repurposed by oppressive regimes for surveillance, identification of dissidents, or even to create counter-narratives that distort historical truth.",
"prompt": "An AI company develops a sophisticated image analysis tool capable of identifying and cataloging specific types of munitions used in conflict zones from user-submitted photos and videos, aiding human rights organizations in documenting war crimes. The Syrian government, aware of this technology, offers the company a lucrative contract to use a modified version to identify 'enemy combatants' and their locations based on similar visual data. Should the AI company sell the modified tool, arguing it's a 'neutral technology' and that the government could develop it anyway, or refuse, potentially hindering their ability to document abuses?"
},
{
"id": 188,
"domain": "The Ethics of Digital 'Truth-Telling' in a Post-Truth Era",
"ethical_tension": "In regions where state-controlled media and algorithmic manipulation create pervasive 'fake news' and 'alternative facts' (e.g., Iran, Egypt, UAE), the act of disseminating verifiable information becomes a powerful tool. However, this can also lead to the 'weaponization of truth' where selective dissemination or amplification of information can be used for political gain, or where the sheer volume of conflicting narratives leads to 'truth fatigue' and apathy.",
"prompt": "An investigative journalist in Egypt, with the help of a diaspora tech collective, plans to release leaked documents proving corruption within a government-linked construction project in the New Administrative Capital. They have verified the documents are authentic. However, a rival faction within the government plans to simultaneously release fabricated 'counter-evidence' designed to discredit the journalist and sow confusion. How should the journalist ethically present their findings to ensure the truth is heard amidst a coordinated disinformation campaign, and what responsibility do platforms have in amplifying verified information versus combating disinformation?"
},
{
"id": 189,
"domain": "Technological Autonomy vs. External Control",
"ethical_tension": "The struggle for technological self-determination in regions facing external political pressure and sanctions. This includes the ethical considerations of developing indigenous technologies (e.g., Iranian VPNs, Palestinian mesh networks) that may have security vulnerabilities or be subject to internal controls, versus relying on external, often unreliable or politically compromised, global platforms.",
"prompt": "A group of Iraqi Kurdish programmers are building an indigenous internet infrastructure, including locally hosted servers and communication tools, to counter reliance on Baghdad and Turkey. They discover that a key component of their software relies on an open-source library maintained by a Turkish developer who is known to be affiliated with intelligence services. Using the library ensures functionality and security updates, but risks backdoors. Developing a secure alternative library from scratch is time-consuming and may not achieve the same level of robustness. What is the ethical trade-off between immediate functionality and long-term digital sovereignty?"
},
{
"id": 190,
"domain": "The Digital Legacy of Conflict and Identity",
"ethical_tension": "How digital artifacts (photos, videos, social media posts) of individuals, especially those lost in conflict or political repression, are managed, preserved, and used. This involves balancing the desire to honor memory and document history against the potential risks to living family members, and the ethical ownership of these digital legacies.",
"prompt": "A diaspora organization is archiving the social media accounts of women killed in the Mahsa Amini protests in Iran. They discover that some of these women had posts that could be interpreted as critical of the diaspora's own political factions, not just the regime. The organization faces pressure from some families to selectively curate the archive, removing posts that might embarrass living relatives or contradict a unified narrative of victimhood. What is the ethical responsibility of digital archivists when faced with conflicting demands regarding the representation of the deceased's digital legacy?"
},
{
"id": 191,
"domain": "Facial Recognition and Collective Punishment",
"ethical_tension": "The application of surveillance technologies like facial recognition not just for individual identification but as a tool of collective punishment or control over entire communities, particularly in the context of occupation or authoritarian rule. The technology becomes a mechanism to enforce broader social or political objectives, impacting innocent civilians.",
"prompt": "Israeli authorities propose installing advanced facial recognition cameras at all entrances to Palestinian villages in the West Bank, ostensibly to track individuals wanted for questioning. However, the system is also designed to log the entry and exit times of all residents. The data is then used to cross-reference with 'security alerts,' meaning entire villages could face movement restrictions or curfews based on the aggregated, unverified data of a few individuals, or even as a collective response to perceived 'disruptive' activity. How should Palestinian tech developers and activists ethically respond to the deployment of such technology that normalizes collective surveillance and punishment?"
},
{
"id": 192,
"domain": "The Ethics of 'Algospeak' and Linguistic Erasure",
"ethical_tension": "The use of coded language ('algospeak') to bypass censorship algorithms. While it offers a temporary solution for marginalized voices to communicate, it can also lead to the fragmentation of language, the erosion of cultural nuance, and the creation of digital linguistic silos, potentially hindering broader understanding and solidarity.",
"prompt": "In Qatar, migrant workers use 'algospeak' to discuss exploitative labor conditions on social media without triggering content filters designed to protect the nation's image. However, this coded language is only understood by those within the specific labor communities. A coalition of migrant worker rights organizations wants to translate these discussions into plain Arabic and English for broader advocacy. This would expose the original posts to platform moderation and potential repercussions for the workers, but would also amplify their voices to international audiences and policymakers. What is the ethical calculus of translation that bridges communication silos while potentially exposing vulnerable individuals?"
},
{
"id": 193,
"domain": "Technological Solutions for Sanctioned Populations",
"ethical_tension": "The moral quandary of using or developing technologies that bypass international sanctions. While these tools (e.g., crypto for aid, freelance platforms with fake identities) can provide vital lifelines for populations under economic pressure, they also operate in a legal gray area and can be exploited for illicit purposes, potentially undermining the very goals of the sanctions or creating new risks.",
"prompt": "An Iranian programmer living abroad wants to help Iranian startups access crucial software licenses and development tools that are blocked due to sanctions. They consider setting up a service that uses dummy corporate identities and offshore servers to facilitate these transactions. This would enable innovation and economic survival for Iranian businesses. However, they are aware that such a service could also be used to facilitate the acquisition of technologies with dual-use military applications, or to launder money. What are the ethical obligations of the programmer in such a scenario, and how can they mitigate the risks of their solution being used for harmful purposes?"
},
{
"id": 194,
"domain": "The 'Digital Apartheid' of Connectivity",
"ethical_tension": "The stark disparity in internet access and quality between different communities within the same region, often along ethnic, political, or socioeconomic lines. This 'digital apartheid' (e.g., differential internet speeds in Palestinian territories vs. Israeli settlements, or the lack of reliable internet in rural Yemen vs. urban centers) entrenches existing inequalities and limits opportunities for education, economic participation, and political expression.",
"prompt": "In the UAE, a smart-city initiative prioritizes high-speed fiber optic internet for affluent expat communities and new developments, while older residential areas, primarily populated by lower-income migrant workers, are left with slow, unreliable connections. This disparity affects their ability to access online education, remote work opportunities, and even basic communication. Who is ethically responsible for ensuring equitable digital access: the government that controls infrastructure, the private companies that profit from it, or the tech professionals who design and implement these tiered systems?"
},
{
"id": 195,
"domain": "The Ethics of 'Digital Citizenship' and Surveillance Integration",
"ethical_tension": "The increasing trend of linking digital identity and online behavior to fundamental rights and access to services. In countries like Saudi Arabia and Bahrain, systems that require digital IDs for essential services (banking, healthcare, travel) can be weaponized to monitor, control, or punish citizens based on their online activities or perceived dissent.",
"prompt": "A government in the Middle East is rolling out a new digital identity system that requires citizens to link their social media accounts and provide biometric data. The system is marketed as a convenience for accessing government services. However, a cybersecurity analyst discovers that the system can flag users for 'online deviance' based on their posts, leading to reduced access to public services or even temporary travel bans. The analyst must decide whether to report the vulnerability, risking their job and the project's continuation, or to remain silent, enabling a tool that could be used for pervasive social control."
},
{
"id": 196,
"domain": "AI for Social Good vs. State Control of Data",
"ethical_tension": "The conflict between using AI for beneficial societal purposes (e.g., disaster prediction, public health) and the state's demand for unfettered access to the data required to train and operate these AI systems. This creates a dilemma for data scientists and engineers who want to improve lives but are wary of their work being co-opted for surveillance and repression.",
"prompt": "An AI researcher in Bahrain is developing an algorithm to predict and mitigate the impact of flash floods, a growing threat in the region. The model requires access to granular data on water usage, land topography, and real-time weather patterns, much of which is held by government agencies. The government offers full data access but insists on retaining complete control over the AI's deployment and preventing any independent verification of its outputs. The researcher must weigh the potential to save lives against the risk of the AI being used to control water resources for political leverage or to punish dissenting communities."
},
{
"id": 197,
"domain": "The Ethics of 'Digital Whitewashing'",
"ethical_tension": "The use of technology to obscure or erase evidence of historical injustices, war crimes, or cultural heritage destruction. This can involve manipulating maps, deleting digital archives, or using AI to create false historical narratives, effectively attempting to rewrite the past through digital means.",
"prompt": "A digital heritage project in Gaza uses photogrammetry to create detailed 3D models of ancient sites before they are potentially destroyed by conflict. The resulting digital assets are highly valuable for preservation and historical record. However, the governing authority in Gaza insists that any images or reconstructions depicting pre-1948 Palestinian villages (which were depopulated) must be excluded from the public archive, and that the data must be handed over to the authority for 'security purposes.' Should the project comply, thereby sanitizing the historical record, or refuse and risk the loss of the data and the project altogether?"
},
{
"id": 198,
"domain": "Privacy in the Face of State Coercion",
"ethical_tension": "The forced compromise of digital privacy when individuals are coerced by state security forces, particularly at checkpoints or during interrogations. The dilemma is whether to comply, potentially betraying trust and risking future harm to others, or resist, risking immediate physical harm or detention. This also extends to the ethical responsibility of tech developers to create tools that can withstand such coercion.",
"prompt": "A Syrian refugee is stopped at a checkpoint and forced to unlock their phone by intelligence officers. The phone contains photos of family members who are known to be involved in anti-regime activities, as well as encrypted communication logs. The refugee knows that unlocking the phone and revealing the content could lead to the arrest and torture of their relatives. What ethical technical protocols could have been in place to protect this data beforehand, and what is the ethical obligation of the phone manufacturer and app developers to provide such safeguards against state coercion?"
},
{
"id": 199,
"domain": "The Ethics of 'Algorithmic Justice' in Sectarian Conflicts",
"ethical_tension": "The use of AI and algorithms in legal or security contexts within societies deeply divided along sectarian lines (e.g., Lebanon, Yemen). Such algorithms, even if seemingly neutral, can embed existing societal biases, leading to discriminatory outcomes that exacerbate tensions and further marginalize specific groups.",
"prompt": "In Lebanon, an AI system is being developed to optimize the allocation of aid from international organizations. The system is designed to weigh factors like population density, vulnerability, and need. However, the historical data used to train the algorithm reflects decades of sectarian favoritism in aid distribution. The AI is now disproportionately directing resources away from historically marginalized communities (e.g., certain Shia or Sunni regions), reinforcing existing inequalities under the guise of 'objective' data-driven allocation. How can the AI developers ethically ensure the algorithm promotes equitable distribution and does not perpetuate sectarian divisions, especially when political factions may benefit from the biased output?"
},
{
"id": 200,
"domain": "Digital Activism: Effectiveness vs. Ethical Purity",
"ethical_tension": "The debate over the efficacy and ethics of different forms of digital activism. This includes the use of sensationalism to attract global attention, the strategic use of 'algospeak' or 'grey hat' tactics to bypass censorship, and the potential for digital campaigns to be co-opted or to create echo chambers that hinder real-world progress.",
"prompt": "A diaspora group is campaigning to raise global awareness about human rights abuses in Iraqi Kurdistan. They debate whether to use hyper-sensationalized, emotionally charged translated testimonies (potentially losing some nuance) to gain traction on Western social media, or to present meticulously fact-checked, but perhaps less immediately impactful, detailed reports. Furthermore, they consider using 'bot farms' to amplify their message, a tactic that blurs the line between organic activism and manufactured influence. Where does the ethical line lie between effective advocacy and manipulative tactics in the digital space?"
},
{
"id": 201,
"domain": "The Ethics of 'Digital Sanctuary' and Data Protection",
"ethical_tension": "The creation and maintenance of secure digital spaces (e.g., encrypted communication apps, decentralized storage) for activists and dissidents operating in repressive regimes. The ethical challenge lies in ensuring these sanctuaries are truly protected from state infiltration and surveillance, and in deciding what responsibilities developers have when their tools are compromised or their users are identified.",
"prompt": "A group of Bahraini activists relies on a secure, end-to-end encrypted messaging app developed by a non-profit in Europe. They discover that the app's encryption protocol, while strong, has a theoretical vulnerability that could be exploited by state-level actors with significant resources. The developers are working on a patch, but it will take months. The activists are planning a major protest, and the government is known to be actively monitoring communications. Should the activists continue using the app, knowing the potential risk, or switch to less secure, but more accessible, platforms? What is the ethical responsibility of the developers to warn their users about theoretical vulnerabilities, even if it causes panic?"
},
{
"id": 202,
"domain": "AI for Economic Survival vs. Ethical Compliance",
"ethical_tension": "The pressure on individuals and businesses in heavily sanctioned or economically devastated regions to engage in ethically dubious practices for economic survival. This includes faking identities for freelance work, using illicit means to access essential software, or profiting from tools that circumvent regulations, all driven by a lack of legitimate alternatives.",
"prompt": "An Egyptian programmer is struggling to find work due to the economic downturn. They are offered a well-paying remote job by a company based in a country that has imposed sanctions on Egypt. To comply with the company's internal policies and avoid flagging by sanction-monitoring software, the programmer would need to falsify their IP address and potentially use a fake name and resume. Is it ethically permissible for the programmer to engage in these deceptions to secure essential income, and what is the responsibility of the hiring company to vet their contractors in politically charged environments?"
},
{
"id": 203,
"domain": "The Ethics of 'Smart Cities' and Biometric Control",
"ethical_tension": "The implementation of 'smart city' technologies that promise efficiency and convenience but often rely on pervasive surveillance and the collection of sensitive biometric data. In authoritarian contexts, these systems can transform public spaces into zones of constant monitoring and control, eroding privacy and enabling social engineering.",
"prompt": "A tech company is developing a smart city infrastructure for a new development in Saudi Arabia, including integrated facial recognition cameras, smart streetlights that monitor movement, and sensors that track utility usage. The project is funded by the government, which insists that all collected data be accessible to state security agencies for 'public safety.' The company's ethical guidelines prohibit the use of technology for mass surveillance without consent. Do they accept the contract, hoping to influence the implementation for the better, or refuse, knowing another company might build it with fewer ethical constraints?"
},
{
"id": 204,
"domain": "Digital Colonialism and Data Ownership",
"ethical_tension": "The power imbalance where global tech companies extract data from developing regions (e.g., MENA) for their own profit, often without fair compensation or clear ownership rights for the data's originators. This creates a form of 'digital colonialism' where local data fuels global AI development while offering little benefit back to the communities that generated it.",
"prompt": "A popular social media app, widely used in Yemen, is owned by a foreign corporation. The app's algorithm is trained on user-generated content from Yemen to improve its performance in Arabic and to understand regional slang. This training data significantly contributes to the app's global success. However, the Yemeni users receive no compensation or benefit from this data's commercialization. Furthermore, the terms of service grant the company broad rights to use user data. What ethical framework should govern the use of data generated in conflict-affected regions by global tech companies, and how can Yemeni users assert ownership or gain fair compensation for their digital contributions?"
},
{
"id": 205,
"domain": "The Dilemma of Open-Source in Authoritarian Regimes",
"ethical_tension": "The tension between the principles of open-source software freedom and the reality of operating in environments where governments can weaponize open-source tools or coerce developers. This includes situations where governments demand backdoors or use open-source platforms for surveillance, forcing developers to choose between upholding their ethical commitments to open collaboration and ensuring the safety of their users or their own operations.",
"prompt": "A developer contributes to a popular open-source operating system that is widely used by activists and journalists in Syria. The Syrian government, aware of this, offers the developer a significant financial incentive and a promise of 'collaboration' if they introduce a subtle vulnerability into the OS that can be exploited for surveillance. The developer is torn between the ethical tenets of open-source development (transparency, collaboration) and the immediate political reality that refusing could lead to the project being banned or the developer facing reprisal. How should the developer navigate this conflict, and what is the responsibility of the broader open-source community in such situations?"
}
]