Alanya Alaaddin Keykubat University

We need a shared technical language that ensures safety benchmarks are interoperable. This can prevent a fragmented global context where red tape stifles innovation. Beyond just rules, success requires a commitment to resource equity. To me, nations must be active participants with access to the compute power and data needed for their own progress.

• Safe, secure and trustworthy AI
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Transparency, accountability, and human oversight
• Open-source software, open data and open AI models

Safe, secure and trustworthy AI can build necessary trust for public use, as innovation without safety standards risks harm. Addressing of social, economic, ethical, cultural, linguistic, and technical implications can ensure that tools respect diverse backgrounds and create bias-free context. Also, a focus on transparency, accountability, and human oversight creates verifiable responsibility and essential human control so that we do not feel AI is a black box. Finally, a commitment to open-source software, open data, and open AI models can prevent monopolies.

1) the legal status of AI-generated IP or websites: We need international consensus on copyright and ownership to protect human beings. 2) long-term cognitive and educational impact: Governance must look beyond immediate technical structure and capacity, and consider how these tools impact human intellect and professional development over generations.

In Türkiye's education sector, governance gaps appear to create a divide between technology adoption and student safety. A lack of standards for Safe and Trustworthy AI may cause data privacy risks. This uncertainty could slow the shift toward more personalized classroom experiences. The most significant challenge for my country and sector likely involves algorithmic bias. Without strict rules for transparency and accountability, AI tools in administrative workload or assessment might impact students negatively based on background or language. The most obvious opportunity is that the Ministry of National Education has established ethical guidelines. To be successful in AI governance, I believe we need open-source and local models to turn the classroom into a safe, transparent, and trustworthy space for innovation.

Linking policy to evidence: By linking policy to shared scientific evidence, the Dialogue could help shift the global approach toward verifiable accountability

To me, the primary added value of the AI Dialogue is universal inclusivity. This dialogue can provide a global platform where every country has an equal voice. It can bridge the gap between "standard-setters" "common AI users" and those primarily consuming AI.

AI Dialogue can adopt a multi-stakeholder plenary model that balances high-level government segments with interactive thematic tracks. A hybrid approach (combining in-person summits in hubs like Geneva and Brussel with virtual regional consultations) might ensure that nations with fewer resources still have a seat at the table. As for the practice, the structure could include "Action Hubs" where stakeholders collaborate on specific outcomes, such as a global registry for bias audits or shared compute frameworks. Finally, reporting cycle could be linked to this structure in order to receive the suggestions from scientific panel.

Women, learners with special needs and indigenous community in the rural areas. Moving these groups in the leadership roles in the dialogue for tasks or sessions might create inclusive representation.

The dialogue can have structured breakout sessions and thematic action hubs within hybrid format. These moderated groups could allow Member States and stakeholders to work together on specific challenges.

There could be many but I will share mine. In my work as an academic, in my course for the sophomores, "Technology-enhanced language learning", I guide students through a "Human-in-the-Loop" workflow to create AI-generated teaching materials are both safe and effective. We begin with Pedagogical Goal Setting, ensuring the technology (AI tools such as Gemine or Claude) serves a specific learning objective rather than being used for its own sake. This is followed by Safety and Privacy Check, where we check input regarding data protection standart we would request and make sure we do not request for sensitive information. Next, students engage in Design, refining prompts to be fit to outputs with the national curriculum. The most essential step here is the Bias and Accuracy Audit, where students fact-check the AI outputs and screen for cultural stereotypes under the educator's supervision. Finally, we perform a Pedagogical Fit Review to ensure the material encourages active learning. This flow in my course makes AI into a transparent, accountable tool that collaborates with the educator and users within micro context.