Hamburg Declaration on Responsible Artificial Intelligence (AI) for the Sustainable Development Goals (SDGs)

Why We Take Action

We are at a crossroads. Despite the progress made in recent years, we need renewed commitment and engagement to advance toward and achieve the Sustainable Development Goals (SDGs). Digital technologies, such as Artificial Intelligence (AI), can play a significant role in this regard. AI presents opportunities and risks in a world of rapid social, political, economic, ecological, and technological shifts. If developed and deployed responsibly, AI can drive sustainable development and benefit society, the economy, and the planet.

Yet, without safeguards throughout the AI value chain, it may widen inequalities within and between countries and contribute to direct harm through inappropriate, illegal, or deliberate misuse. It can also contribute to human rights violations, fuel disinformation, homogenize creative and cultural expression, and harm the environment. These risks are likely to disproportionately affect low-income countries, vulnerable groups, and future generations. Geopolitical competition and market dependencies further amplify these risks.

A Joint Vision

As global stakeholders from governments, international organizations, the private sector, academia, and civil society, we convene in Hamburg to shape a human-centric, human-rights-based, inclusive, open, sustainable, and responsible AI future. We commit to advancing AI for the SDGs, aligning with ongoing international efforts[1].

A responsible AI future must be built on equal and meaningful participation, with actions to ensure that all stakeholders, especially those from emerging markets, developing economies, and vulnerable groups, have fair and equitable access to, as well as ownership of, computing, data, investment, and resources for capacity and talent development. AI’s benefits must not remain concentrated among a privileged few.We are committed to bridging digital divides and empowering all nations and communities to co-create and leverage AI solutions and evaluations that serve people and the planet.

To achieve this, we call for leveraging AI responsibly, inclusively, and sustainably, aligned with the five pillars of the 2030 Agenda for Sustainable Development: people, planet, prosperity, peace, and partnerships.

People – Advancing Human Rights, Gender Equality and Inclusivity

We will design, develop, and deploy AI systems that respect, protect, and promote human rights, inclusion, gender equality, human dignity, and safety. We aim to mitigate harm and ensure that the benefits of AI are accessible to everyone, particularly in emerging markets and developing economies. We will strive for the meaningful and active participation of stakeholders, particularly marginalized and vulnerable groups, in AI design, development, deployment, and governance.

• Human-Rights-Based and Human-Centered AI   
  The promotion, protection, and upholding of human rights must be applied throughout the design, development, and deployment lifecycle and value chain of AI systems and related data usage, ensuring fairness, transparency, and accountability mechanisms for all actors involved. AI systems should be non-discriminatory and privacy-preserving, respect data free flow with trust, promote cultural diversity, freedom of opinion and expression, and freedom of religion or belief. AI systems should have mechanisms for mitigating harmful algorithmic biases and outcomes, which might perpetuate inequalities and impact, in particular, children, women, and girls, as well as marginalized groups.    
  
  
• Equity and Inclusive Participation 
  AI systems should be accessible and beneficial to all, including as a digital public good when applicable, with an emphasis on the inclusion of women, girls, children, youth, Indigenous Peoples, persons with disabilities, marginalized communities, individuals belonging to minority groups, and underserved populations. Accessibility should be enhanced through AI literacy and public consultations when designing large-scale, public-facing AI systems. Participation in AI development and workforce diversity, particularly in top-tier research and AI development roles, should be strengthened by creating equitable and local economic opportunities across all regions.        
  
  
• Trust, Safety, and Protection from Harm  
  AI and related data usage should be designed, developed, and deployed with robust safeguards to ensure privacy, safety, integrity, transparency, and dignity while preventing exploitation and social and societal harms. These safeguards should be consistent and include governance frameworks consistent with international law, including international human rights and humanitarian law, along with clear accountability and redress mechanisms. They should also incorporate due diligence mechanisms, including testing and auditing, with inputs and ownership from local communities where AI solutions are deployed. Stakeholders should conduct risk assessments to anticipate and mitigate any potential misuse of AI systems.

Planet – Sustainable AI for Climate and Environmental Stewardship

We will align AI usage with global climate and environmental goals, including those outlined in the Paris Agreement. We aim to mitigate the impacts of AI and accelerate AI-driven solutions that augment our capacities to advance global sustainability and address climate challenges while contributing to ecological balance.

• Green Compute and AI       
  AI systems should be designed, developed, and deployed to maximize energy efficiency and environmental sustainability throughout their value chain and lifecycle. These efforts should include optimizing energy needs for computing, reducing carbon and material footprints, reusing and recycling e-waste, and promoting the use of sustainable energy and efficient water usage in AI infrastructure, particularly in networks and data centers.           
  
  
• AI for Climate Action, Advancing Sustainability and Environmental Solutions    
  AI should accelerate solutions for climate change mitigation and adaptation, biodiversity conservation, sustainable supply chains, and circular economies. Through targeted capacity-building programmes and support for Research & Development and infrastructure, developers, particularly in emerging markets and developing economies, should be empowered to design and deploy climate-focused AI innovations locally.         
  
  
• Transparency and Accountability in AI’s Environmental Impact
  Stakeholders should leverage robust standards and frameworks to assess AI’s environmental footprint and associated risks[2], including the proportionality of environmental impact across the AI value chain and life cycle. Policymakers and industries, in collaboration with other stakeholders, should work together to mitigate AI’s environmental footprint, promote transparency, and ensure that AI development aligns with global climate and sustainability goals.

Prosperity – Economic Growth, Innovation, and Equity

We will strive to ensure that AI systems support inclusive economic and social development by creating opportunities that foster local innovation and minimize economic divides across and within countries.

• Sustainable AI and Data Infrastructure to Support Economic Growth          
  AI-driven prosperity shouldbe built on inclusive, accessible, and sustainable AI and data infrastructure and tools. This includes investments in connectivity infrastructure, computing resources, and AI as digital public goods when possible, as well as support for underserved and under-resourced language groups in AI training, AI-ready digital infrastructure, and interoperable platforms.       
  
  
• AI as an Engine for Economic Growth and Local Innovation           
  Local AI ecosystems, entrepreneurs, and small and medium enterprises, including startups, should be supported with funding and mentorship to enhance economic resilience and drive industrial transformation. Local and diverse languages should be represented and included in AI development and deployment through mechanisms such as licensing models that maximize societal benefits and respect the rights and incentives of data subjects and owners.
  
  
• Skills Development and Workforce Readiness    
  To fully harness AI’s potential and manage workforce transition, citizens and employees should be equipped with AI-related skills through education, training, and upskilling initiatives. Special emphasis should be placed on technical capacity building, fostering interdisciplinary scientific education, and enhancing digital literacy, ensuring that all economies can participate in, shape, and benefit from the AI-driven future.

Peace – Integrity and Cohesive Societies

Acknowledging the close relationship between peace and sustainable development, we will strive to ensure AI systems are not misused to disrupt peace and undermine the rule of law and governance, especially through large-scale information manipulation, including misinformation and disinformation campaigns. We aim to ensure that AI systems are secure, maintain public confidence, and are not used against the planet and people.

• Preserving and Promoting Information Integrity: Through robust safeguards and methods, such as provenance tracking, AI systems should be designed, developed, and deployed to protect the integrity of information ecosystems and combat disinformation, deepfakes, and algorithmic manipulation, particularly in critical areas like news, healthcare, judicial systems, and elections.             
  
  
• Safeguarding Public Trust and Promoting Safe Digital Spaces: AI systems should foster cooperation, public trust, and societal cohesion while mitigating the amplification of harmful rhetoric, division, or societal disruption, including online violence, particularly against children, women, and girls, as well as marginalized groups.

Partnership – Deepening Global Cooperation for Responsible AI

We are determined to strengthen multistakeholder international partnerships and foster a globally inclusive AI ecosystem that upholds equity, innovation opportunities, and shared responsibility, aligned with global sustainable development objectives.  

• Multistakeholder Collaboration: New and existing multistakeholder partnerships should ensure that all voices are heard, particularly those from emerging markets and developing economies. They should also connect with existing initiatives to pool resources, such as infrastructure, computing, models, and data sets, for driving AI innovation and scaling up AI capacity-building.
  
  
• Open Access and AI as Digital Public Goods: To foster trust and accelerate innovation, particularly in critical areas such as climate change mitigation, science, and healthcare, where feasible, AI systems and associated datasets (publicly available, non-personal, and anonymized) should be treated as global digital public goods. They should include robust evaluation mechanisms, be interoperable and open-sourced, adhere to international standards, and be available under standard open licenses.
  
  
• Knowledge sharing and mutual learning: Stakeholders should facilitate and fund interdisciplinary research, promote scientific cooperation, and pool knowledge resources to foster mutual learning and drive the interoperability of global standards and approaches. 

We invite stakeholders who share our values and principles to join us in advancing the design, development, and deployment of responsible, safe, secure, and trustworthy AI globally, with the goal of achieving the SDGs.

Together, we will drive collective global action in line with the above and agree to stay connected through stakeholder engagement, tracking progress, and exploring joint partnerships, collaboration, and activities to achieve our joint vision.

[List of Endorsements]

Annex

Commitments

While the Declaration and the following voluntary general commitments are not legally binding, stakeholders are highly encouraged to endorse them – as a collective affirmation.

To deliver on the Declaration’s priorities:

General Commitments

1. People: We commit to advancing inclusive and responsible AI in alignment with international human rights frameworks, safeguarding human dignity and equality.
2. Planet: We commit to enhancing the environmental sustainability of AI by improving its resource efficiency, promoting the use of sustainable and renewable energy, minimizing carbon emissions across its lifecycle, and leveraging AI applications and data to support climate and environmental action.
3. Prosperity: We commit to advancing AI as a catalyst for inclusive economic growth by supporting local innovation and developing AI ecosystems across all regions. We further commit to expanding equitable access to AI education and skills development, focusing on women, girls, children, and marginalized communities.
4. Peace: We commit to implementing robust safeguards against the misuse of AI to uphold information integrity and advancing multilevel cooperation and media literacy as key enablers of secure, transparent, and accountable AI systems.
5. Partnerships: We commit to advancing global collaboration through inclusive partnerships that enable the responsible sharing of data and AI solutions as digital public goods, where appropriate, to address global challenges.

Additional Individual Commitments

This optional section gives the opportunity for endorsers to add their individual targets and commitments, which they would like to highlight in this section.

BMZ:

• We commit to sharing 12 AI building blocks for climate action globally as global digital public goods by 2026 for replication (data-sets and AI models).
• We commit to training and building capacities for 160,000 people in AI across the global majority countries by 2030.
• We commit to share globally 2 proven safeguard mechanisms developed by policy maker communities and practitioners in Global Majority countries, namely the “AI policy Playbook” and the “responsible AI assessments”.
• We commit to making 31 open AI datasets available as digital public goods by 2026 (focusing on the following SDGs: Zero Hunger, Affordable and Clean Energy, Climate Action, Life on Land, Reduced Inequalities).
• We commit to developing, disseminating, and maintaining, through multistakeholder cooperation, 45 sets of safe and secure open-source software, open data, open AI models, and open standards that benefit society as a whole.

UNDP:

• We commit to supporting 15 countries in their AI transformation journey by leveraging the AI Landscape Assessment framework (AILA). 
• We commit to supporting 10 countries in the local language digitalization and improving the representation of underrepresented languages in AI design, development, and deployment. 
• We commit to training knowledge workers, comprising 25% of the UNDP workforce, to have a baseline knowledge of working with AI, and to train over 300 senior managers on leadership in the age of AI.
• We commit to leveraging a combination of UNDP’s open-source software, reusable patterns, design guidance, advisory, and technical support to directly or indirectly:
  • Support 20 countries by 2030 in establishing digital carbon registries for carbon markets.
  • Support 15 countries by 2030 in developing national climate transparency platforms as digital public infrastructure.
• We commit to designing a tailored Digital Readiness Methodology for the Biodiversity and Ecosystems Sector, building on UNDP’s existing Digital Readiness Assessment (DRA) framework. This methodology will provide countries with a practical tool to evaluate their readiness and identify opportunities for digital transformation in support of biodiversity goals.

[1] The World Summit on Information Society (WSIS) framework, the United Nations (UN) Global Digital Compact, UNESCO Recommendation on the Ethics of AI, The UNGA Resolution A/RES/78/265, Outcomes of the February Paris AI Action Summit, The Bletchley Declaration, The Africa Declaration on Artificial Intelligence and other efforts.

[2] This can include energy and water consumption, greenhouse gas emissions, e-waste and other environmental risks.

Stay ahead in the world of AI, business, and technology by visiting Impact AI News for the latest news and insights that drive global change.