RamSmith & Associates

As a patent agent and advocate working at the intersection of artificial intelligence and intellectual property law, I believe the first Global Dialogue on AI Governance would be a success if it achieves three concrete outcomes. First, the Dialogue must produce a clear, actionable Co‑Chairs' Summary that addresses interoperability of AI governance approaches across jurisdictions. This summary should include practical recommendations on transparency, accountability, and human oversight of AI systems, with specific attention to computer‑related inventions and the treatment of AI‑generated outputs under existing intellectual property frameworks. Second, the Dialogue should establish concrete capacity‑building mechanisms for developing countries. These mechanisms must go beyond general statements and include timelines and responsible entities for bridging AI divides, particularly in access to high‑performance computing, open data, and open‑source AI models. Third, I would consider the Dialogue a success if it agrees on a lightweight continuity mechanism beyond the inaugural meeting, such as thematic follow‑up tracks on legal and ethical implications of AI. This would ensure that the discussions on human rights, liability, and dispute resolution for AI‑related disputes do not end with the July 2026 meeting but evolve into sustained multi‑stakeholder cooperation.

• Safe, secure and trustworthy AI
• AI capacity-building
• Interoperability of governance approaches

Safe, secure and trustworthy AI: Essential for building legal and technical confidence in AI systems, particularly in regulated sectors such as medical devices, autonomous vehicles, and industrial automation. AI capacity-building: Developing countries like India face significant gaps in high-performance computing, skilled personnel, and access to AI tools. Concrete mechanisms to bridge these divides are a near-term necessity. Interoperability of governance approaches: As a practitioner handling patent filings across multiple jurisdictions, I see firsthand the need for compatible legal frameworks on AI-related inventions, data, and liability to avoid fragmentation. Transparency, accountability, and human oversight: Critical for ensuring that AI systems comply with existing IP and contract laws, and for designing dispute resolution mechanisms (including ADR) for AI-generated outputs and model training disputes.

From my perspective as a registered patent agent and a legal professional with technical experience in AI, I believe the proposed thematic clusters are largely comprehensive. However, I see two cross-cutting issues that would benefit from more explicit attention. First, intellectual property rights and AI training data, including copyright, fair use, and patentability of AI-generated inventions, are not directly addressed. Clear IP frameworks are essential for fostering innovation, ensuring equitable benefit-sharing, and avoiding legal uncertainty, particularly for developers and startups in developing countries. Second, the environmental sustainability of AI systems (energy consumption, water usage for cooling data centres, and electronic waste) is an emerging governance gap. While the Social Development Goals include climate action, the draft themes could more explicitly link AI governance to ecological footprints and green AI practices. Both issues are cross-cutting, touching on human rights, capacity-building, and technical interoperability. I would encourage the Co-Chairs to consider including these dimensions in the thematic discussions or in the Co-Chairs' Summary, as they are increasingly relevant to industry, policymakers, and civil society alike.

From my perspective in India's intellectual property and industrial automation sector, the two governance gaps I highlighted, IP rights for AI training data and environmental sustainability of AI, have tangible effects. Challenges: On IP, India lacks clear guidelines on whether AI-generated inventions are patentable or who owns copyright over training data outputs. This creates legal uncertainty for our patent filings, especially for computer‑related inventions. Many clients hesitate to invest in AI‑driven R&D due to unclear ownership and infringement risks. On the environmental front, India is witnessing rapid data centre expansion. Without AI‑specific energy and water efficiency norms, we risk worsening water stress in regions like Gujarat (where I am based) and increasing e‑waste from accelerated hardware cycles. Opportunities: These gaps also open doors. India can help shape a pro‑innovation, flexible IP framework for AI that balances creators' rights with access to training data, especially for resource‑constrained developers. The Global Dialogue offers a platform to propose solutions like mandatory disclosure of AI training data sources or a voluntary code on green AI computing. On sustainability, India can become a leader in low‑energy AI inference and cooling technologies, leveraging our growing renewable energy capacity. Aligning AI governance with climate action would not only reduce harm but also create exportable green tech solutions. I see a real chance for India to move from being a rule‑taker to a rule‑shaper in AI governance, building capacity while protecting our environmental and innovation interests.

The Global Dialogue on AI Governance can play a critically constructive role in advancing international cooperation by addressing precisely the cross-cutting gaps I identified. First, the Dialogue can catalyse normative convergence on IP and training data. No single country can resolve questions of copyright, fair use, or inventorship for AI-generated outputs alone. The Dialogue can facilitate a multi‑stakeholder process to develop shared principles or a model framework, reducing legal fragmentation and giving developing countries a voice in shaping rules that affect their innovators. Second, the Dialogue can drive measurable cooperation on AI environmental sustainability. It could launch a voluntary reporting mechanism for energy and water use in AI model training and deployment, alongside a collaborative platform for sharing low‑carbon computing best practices. This would turn sustainability from an abstract concern into actionable, trackable commitments. Beyond these, the Dialogue's unique value lies in its inclusivity. By bringing together member states, technical experts, civil society, and private sector actors – including patent agents and industry professionals from the Global South – it can translate diverse perspectives into tangible outcomes, such as the Co‑Chairs' Summary, thematic follow‑up tracks, or a continuity platform for action. This prevents AI governance from being shaped solely by a few powerful actors. In summary, the Dialogue can act as a bridge‑builder: harmonising norms, enabling collective action on emerging risks, and ensuring that capacity‑building reaches those who need it most. That is the kind of practical, structured cooperation I hope to see emerge from the July meeting.

As a registered patent agent and advocate with experience in computer-related inventions and AI dispute resolution, I see clear opportunities for the Global Dialogue on AI Governance to build upon several existing initiatives. First, the Dialogue should connect with WIPO's work on AI and intellectual property, including the Patent Cooperation Treaty system and the WIPO Arbitration and Mediation Centre's ADR procedures for AI model training disputes. These provide practical frameworks that can be extended and harmonised. Second, it should leverage the AI for Good Global Summit, the UN's Independent International Scientific Panel on AI, and capacity-building programmes such as those offered by the ITU and UNESCO. These mechanisms already support multi-stakeholder collaboration and digital inclusion. Third, the Dialogue should engage with open-source software and open-data initiatives, including open AI models, to ensure transparency and interoperability. The added value of this Global Dialogue lies in its mandate to bridge existing silos. It can provide a coherent platform where IP frameworks, technical standards, capacity-building efforts, and human rights considerations are discussed together. Specifically, it can propose practical solutions for addressing copyright and patent issues arising from AI training data, promote harmonised guidelines for computer-related inventions across jurisdictions, and ensure that developing countries gain access to high-performance computing skills and AI applications. By doing so, the Dialogue can create a continuity mechanism that turns fragmented efforts into a concerted global strategy for safe, secure, and trustworthy AI governance.

As a registered Indian patent agent, an advocate, and a professional with over 14 years of industry experience in AI, semiconductors, and IoT, I see multiple ways for stakeholders to contribute meaningfully to the Global Dialogue. First, stakeholders can provide written inputs through the UN web platform, as I have done previously for Indian patent office guidelines and WIPO competitions. Second, they can participate in virtual consultations like this one, offering technical and legal perspectives on the four thematic clusters – particularly on open-source AI models, transparency, and human rights. Regarding the format and structure, I recommend three improvements. First, allocate dedicated breakout sessions for intellectual property and computer-related inventions, as these are critical for trustworthy AI systems but often underrepresented. Second, ensure that capacity-building for developing countries includes practical mechanisms, such as pro bono legal and technical support, to bridge AI divides. Third, establish continuity mechanisms beyond the inaugural Dialogue, ideally a standing working group on AI and intellectual property interoperability, given the rapid evolution of open-source models and cross-border data governance. These recommendations build on my prior submissions to WIPO and the Indian Patent Office, and I am confident they would make the Dialogue more inclusive, actionable, and forward-looking.

Based on my experience as a registered patent agent and an engineer working in industrial automation and power plant operations in India, I see that several voices remain underrepresented in global AI governance discussions. First, practitioners from developing countries who handle computer-related inventions (CRIs) and AI patents daily are rarely consulted. Their practical insights on how AI systems are built, deployed, and litigated at the local level could inform more realistic governance frameworks. Second, industrial workers and mid-level engineers in manufacturing sectors, where AI is increasingly deployed, have limited platforms to share their perspectives on safety, accountability, and human oversight. Their on‑the‑ground experience with AI‑driven control systems and automation matters for designing meaningful human oversight mechanisms. Third, smaller enterprises and individual innovators from non‑OECD countries are often absent. They face unique capacity gaps in high‑performance computing and open‑source AI models, yet their needs are seldom reflected in policy drafts. To include these voices, I suggest the following practical steps: 1. Create dedicated virtual consultation slots with flexible timing and interpretation in multiple languages, building on the inclusive approach already adopted by the Co‑Chairs. 2. Establish a structured channel for written submissions from patent agents, industry engineers, and small business representatives, with simplified templates and targeted outreach. 3. Offer mentorship or fellowship opportunities to practitioners from developing countries to participate actively in thematic tracks of the Global Dialogue. 4. Recognise technical‑legal expertise as a distinct stakeholder category, ensuring that individuals like patent agents and in‑house engineers are invited explicitly, not just through broad civil society categories. These steps would make the Dialogue more transparent, substantive, and truly global.

Based on my experience as a patent agent and legal professional working at the intersection of AI, IP, and industrial compliance, I believe the Global Dialogue would benefit from three innovative engagement formats. First, I recommend structured breakout sessions organised around specific problem statements tied to the four thematic clusters. These sessions should produce tangible outputs such as model clauses for interoperability or shared principles for open-source AI models. Second, a "lightning consultation" format would be effective. In this model, stakeholders deliver three-minute pitches on concrete solutions, followed by five minutes of focused Q&A from peers. This mirrors the rigorous time limits of patent examination and legal drafting, ensuring precision and relevance. Third, I see strong value in a collaborative drafting room. Using a shared digital whiteboard, participants could co-develop recommendations on capacity-building mechanisms or continuity platforms in real time. This transforms passive listening into active co-creation. All three formats should feed directly into the Co-Chairs' Summary. A designated rapporteur for each thematic cluster would ensure stakeholder inputs are captured verbatim and attributed. These engagement methods respect the three-minute intervention limit while enabling deeper, solution-oriented exchange.

Based on my experience as a registered Indian patent agent and a legal professional with a technical background in AI, I would like to share three concrete examples. First, the draft guidelines for examining Computer-Related Inventions (CRIs) issued by the Indian Patent Office provide a practical policy approach. They clarify how patent law applies to AI and software, balancing innovation with legal certainty. This reduces ambiguity for inventors and examiners. Second, the WIPO Arbitration and Mediation Centre offers a platform for resolving IP disputes arising from AI model training. As a member of WIPO ADR Young, I have seen how tailored alternative dispute resolution mechanisms can address challenges related to copyright and data ownership without costly litigation. Third, the open call for written submissions by this Global Dialogue, available in all six UN languages, is an inclusive practice. It allows stakeholders from developing countries to contribute meaningfully to AI governance discussions. These examples show that clear legal frameworks, accessible dispute resolution platforms, and inclusive consultation mechanisms offer actionable solutions. I encourage the Co-Chairs to consider these as models for promoting transparency, accountability, and capacity-building in AI governance.