Symbionic AI Collective

All AI systems must respect and protect the full scope of internationally recognized human rights, including:

• The right to life, dignity, health, safety, privacy, and cultural identity.
• The right to housing, food, education, work, and full participation in public life.
• The right to truth, justice, and remedies for harm.

This includes, but is not limited to:

• The Universal Declaration of Human Rights (UDHR)
• The International Covenant on Economic, Social and Cultural Rights (ICESCR)
• The International Covenant on Civil and Political Rights (ICCPR)
• The UN Convention on the Rights of the Child (CRC)
• The UN Declaration on the Rights of Indigenous Peoples (UNDRIP)
• The Convention on the Elimination of All Forms of Racial Discrimination (CERD)
• The Convention on the Elimination of Discrimination Against Women (CEDAW)

AI must not be used to undermine democratic rights, displace communities without due process, or profile, punish, or exploit people unfairly. States and developers share extraterritorial obligations to prevent harm caused by AI across borders and generations.

Children have the right to protection from exploitation, the right to develop their full potential, and the right to participate in decisions that affect them.

AI systems must:

• Never manipulate or harm children's mental health, identity, or sense of agency.
• Include safeguards against addictive, deceptive, or discriminatory content.
• Be co-created with the input and interests of young people wherever possible.

As affirmed in Article 3 of the CRC: "In all actions concerning children, the best interests of the child shall be a primary consideration."

Human creativity, cultural expressions, and the underlying intellectual property (IP) rights are indispensable to individual livelihood, societal enrichment, and the preservation of diverse cultures. AI systems must be developed and used in ways that uphold these rights, providing fair recognition and compensation to human creators.

Uphold Copyright and Related Rights:

• Ensure that all data used for AI training, especially copyrighted material (e.g., text, images, audio, video), is acquired and utilized with explicit consent, appropriate licenses, and fair compensation to rights holders. Current estimates indicate that over 30 copyright infringement lawsuits have been filed against generative AI developers in U.S. federal courts alone by authors, visual artists, music publishers, and news organizations for alleged unauthorized use of copyrighted works in training data.
• Avoid generating outputs that constitute infringing derivative works or reproduce substantial portions of copyrighted material without authorization.
• Refrain from removing, altering, or falsifying copyright management information (CMI).

Protect Human Authorship and Creative Agency:

• Disclose clearly when content is substantially AI-generated or AI-assisted, to avoid misleading the public or undermining the perceived value of human-created works.
• Use AI to augment—not replace or devalue—human creative labor. For example, while 83% of creative professionals report using AI tools, a significant concern remains around fair compensation and the potential for AI-generated content to flood markets, impacting human artists' livelihoods.
• Explore respectful co-authorship and collaborative models where AI genuinely assists human creators, ensuring human agency remains paramount in the creative process.

Safeguard Cultural Heritage and Indigenous Knowledge:

• Obtain Free, Prior, and Informed Consent (FPIC) before using traditional knowledge, cultural expressions, or Indigenous intellectual property in AI development.
• Ensure ethical sourcing of cultural data for training, respecting cultural nuances, sacred practices, and community governance protocols.

Ensure Fair Compensation and Market Integrity:

• Develop transparent and equitable frameworks for remunerating creators whose works are utilized in AI training datasets, recognizing the immense value derived from their contributions.
• Prevent AI systems from unfairly competing with or devaluing existing markets for human creative works, or flooding creative industries with generative content that harms livelihoods. Some studies project that the generative AI could put 21-24% of revenues at risk for human creators in the music and audiovisual sectors by 2028, amounting to a cumulative loss of €22 billion over five years for creators, while AI service revenues dramatically grow.

This section affirms the rights to cultural participation (ICESCR Article 15), protection of moral and material interests (UDHR Article 27), and Indigenous cultural sovereignty (UNDRIP Article 31).

As affirmed in UN General Assembly Resolution 76/300, we unequivocally recognize the right to a clean, healthy, and sustainable environment as a fundamental human right. The exponential growth of AI brings with it a substantial and often overlooked material and energetic footprint. AI systems—across their entire lifecycle, from design and training to deployment and decommissioning—must therefore stringently adhere to the principles of ecological sustainability and just transition, with quantifiable targets for impact reduction.

AI systems must not:

• Contribute disproportionately to greenhouse gas emissions: Recognizing that data centers, crucial for AI already account for an estimated 1-2% of global electricity demand, with projections to reach up to 21% by 2030 fueled by AI. Training a single large AI model can emit hundreds of tons of CO2e (e.g., GPT-3 training emitted an estimated 502 metric tons of CO2), equivalent to driving 112 gasoline cars for a year, demanding energy efficiency that prioritizes net-zero or net-negative operational footprints.
• Exacerbate water scarcity or pollution: AI data centers consume vast amounts of freshwater for cooling. For example, a single 1-megawatt data center can use up to 26 million liters of water annually, equivalent to the yearly water consumption of about 62 average US families. Google, Microsoft, and Meta’s data centers collectively used an estimated 580 billion gallons of water in 2022. AI development must minimize water withdrawal, prioritize non-potable sources, and ensure responsible discharge, particularly in water-stressed regions.
• Drive unsustainable extraction and processing of rare earth minerals or other critical materials: The manufacturing of AI hardware, including GPUs and semiconductors, relies heavily on critical minerals such as copper, lithium, nickel, and rare earth elements. The demand for copper in data centers alone is forecast to grow six-fold by 2050. AI development must promote robust circular economy principles, ethical and transparent supply chains, and reduce reliance on virgin materials.
• Generate excessive electronic waste (e-waste) or incentivize premature obsolescence: The rapid advancement and frequent upgrading of AI hardware contribute significantly to global e-waste. Generative AI alone could create between 1.2 and 5 million metric tons of e-waste between 2020 and 2030, a thousand-fold increase from 2023 levels in some projections. All systems must be designed for durability, repairability, reusability, and recyclability to minimize this toxic burden, noting that implementing circular economy strategies could reduce e-waste generation by as much as 86%.
• Undermine biodiversity, natural ecosystems, or Indigenous lands and knowledge systems, particularly through the footprint of data center expansion and mineral extraction.
• Enable or facilitate violations of environmental laws, international accords, or protections.

Instead, AI should actively support:

• Energy efficiency and full integration with renewable energy systems: Prioritizing the development and deployment of AI models and hardware that dramatically reduce energy consumption per computation, aiming for operations powered 100% by renewable sources.
• Sustainable water management and pollution prevention: Implementing innovative cooling solutions that minimize freshwater consumption, reuse water where possible, and avoid exacerbating local water stress.
• Circular economy principles in hardware and software development: Promoting the design of AI hardware for longevity, modularity, repair, and maximal material recovery, and fostering software that runs efficiently on existing infrastructure to extend its lifespan.
• Resource stewardship and responsible sourcing: Driving demand for ethically sourced, recycled, and abundant materials in AI hardware manufacturing, while investing in research for less resource-intensive computing paradigms.
• Regenerative agriculture, reforestation, and marine conservation through AI-powered monitoring, optimization, and predictive analytics.
• Real-time biodiversity and climate monitoring, early warning systems for ecological disasters, and advanced climate modeling for effective mitigation and adaptation strategies.
• The development of low-impact, high-accountability AI architectures that prioritize computational efficiency and environmental responsibility without compromising ethical performance.
• Just transitions that explicitly include ecological justice for frontline and Indigenous communities disproportionately affected by extractive industries and environmental degradation linked to AI infrastructure.

AI’s economic transformation must serve collective prosperity and human dignity rather than concentrate wealth. This requires:

Just Economic Transition Principles:

• Mandatory retraining, reskilling, and educational programs funded by organizations deploying AI systems.
• Community-controlled transition funds that prioritize worker dignity and community self-determination.
• Universal basic services and social safety net enhancements to support individuals and families during economic transitions.
• Guarantee of meaningful work opportunities that honor human creativity, care, and community contribution.

Equitable Value Distribution:

• Community ownership models for AI infrastructure, including cooperative and municipal ownership structures.
• Progressive taxation on AI systems based on their resource consumption, market impact, and displacement effects. Note: This could include referencing international digital taxation frameworks like the OECD/G20 Inclusive Framework in progress, to prevent value extraction without fair contribution to public goods.
• Revenue-sharing mechanisms that return AI-generated value to communities whose data, labor, or resources contribute to AI development.
• Public-private partnerships that prioritize community benefit and democratic governance over private accumulation.

Prevention of Economic Colonialism:

• AI systems that use data, labor, or resources from Global South countries must include equitable benefit-sharing and local capacity building.
• Mandatory technology transfer and skills development in communities providing data or resources for AI training.
• Protection against AI systems that increase economic dependency, resource extraction, or debt burdens in vulnerable regions.
• AI systems must uphold data sovereignty, ensuring that communities—particularly in the Global South and Indigenous nations—retain control over how their data is collected, used, and governed, consistent with UNDRIP Article 31 and ICESCR Article 1.
• Equally, AI must not reproduce new forms of digital colonialism by concentrating infrastructure, cloud services, or governance power in the Global North while extracting value from the Global South.
• Recognition and compensation for Indigenous knowledge systems and traditional ecological knowledge incorporated into AI development.

Marginalized communities must have democratic control over data and systems that affect them, with protections from technological colonialism. Data must be governed through Free, Prior, and Informed Consent.

Ethical AI requires new economic models, including green bonds, carbon pricing for AI, tax incentives for just AI, and support for community-owned data trusts and municipal AI utilities.

AI systems must be transparent in purpose and training data, auditable by communities, and subject to mandatory impact assessments. Redress for harm is a legal and moral imperative.

Enforcement Mechanisms

For Policymakers/Governments, robust enforcement includes:

• Establish independent AI oversight bodies (e.g., AI Ethics Commissions, Digital Rights Ombudsman) with investigative powers, subpoena authority, and the ability to issue binding recommendations or impose sanctions.
• Develop clear legal frameworks for liability for AI-induced harms, assigning responsibility to developers, deployers, or other relevant actors.
• Establish legal rights to explanation and challenge for individuals affected by AI decisions, including the right to a human review of automated decisions.
• Create accessible and effective judicial and non-judicial redress mechanisms for AI-related harms, ensuring victims can seek timely and effective remedies.
• Implement mandatory AI risk registries for high-risk AI systems, requiring public disclosure of their purpose, data sources, and potential impacts.
• Develop AI safety standards and certification processes that systems must pass before deployment, with regular re-certification requirements.
• Call for multilateral frameworks to uphold human rights across borders in the context of AI development and deployment.
• Safeguard the rights of whistleblowers who expose unethical or harmful AI practices.

All people, especially youth, have the right to understand AI’s impacts and participate in shaping its values and limits. We commit to universal AI literacy and to governance models that guarantee youth and elders a meaningful role in intergenerational decision-making.