AI SISTERS - KENYA

Right now, the quality, safety, and fairness of AI outputs heavily depend on how people prompt. Yet this skill is uneven, undocumented, and largely ungoverned. If the dialogue addressed this, success would look like: Agreed guidelines on responsible prompting (e.g., avoiding harmful, biased, or manipulative instructions) Public education frameworks So prompt literacy becomes a core digital skill like basic computer literacy Workplace standards Organizations training staff on secure prompting (especially to prevent data leakage or Shadow AI risks) Transparency in prompt–output relationships Helping users understand how inputs shape outcomes (reducing blind trust in AI) Equity in AI usage Because better prompting = better results → without this, AI advantages remain with a few The hidden truth: We focus so much on regulating AI systems… but the human layer (prompting) is the control surface. If we ignore it, we leave a major gap in governance.

• AI capacity-building
• Safe, secure and trustworthy AI

AI Capacity-Building -What it really means AI capacity-building is about equipping people, organizations, and countries with the skills, knowledge, and tools to use AI effectively and responsibly. It goes beyond simply having access to AI - it ensures users truly understand how to apply it well. This includes developing practical skills, raising awareness of risks like bias and data leakage, improving access so regions like Africa are not left behind, and strengthening institutional readiness through policies, governance, and training. Why this is urgent Today, many people are using AI blindly, often without guidelines, increasing risks such as misuse and Shadow AI in organizations. At the same time, there is a growing gap between those who build AI and those who use it. Without deliberate capacity-building, this gap becomes dangerous-limiting benefits, increasing vulnerabilities, and slowing responsible adoption.

Yes, one key cross-cutting issue often overlooked is prompt literacy, or how people interact with AI systems. While governance focuses on how AI is built, the quality and safety of outputs depend heavily on user inputs. Poor prompting can lead to bias, misinformation, or data leakage, especially in workplaces. Strengthening prompt literacy through training and awareness is essential to ensure safe, equitable, and effective use of AI.

Governance gaps in prompt literacy are affecting our country and sector as AI adoption grows faster than user awareness. Many people use AI without proper guidance, leading to risks such as misinformation, biased outputs, data leakage, and the rise of Shadow AI in organizations. At the same time, this presents an opportunity to integrate prompt literacy into training, policies, and capacity-building efforts. Strengthening how users interact with AI will reduce risks, improve outcomes, and support more responsible and inclusive adoption.

The AI Dialogue can advance international cooperation by making prompt literacy a global priority. It can help establish common guidelines for responsible prompting, reducing risks like bias, misinformation, and data leaks. The Dialogue can also promote collaborative capacity-building, enabling countries to share training, tools, and best practices to strengthen how users interact with AI. By embedding prompt literacy into governance frameworks, it ensures AI is safely and responsibly used—not just well-regulated at the system level, but effectively guided by human users

Existing initiatives like AI SISTERS, regional AI hubs, and global programs by UNESCO and the World Economic Forum focus on AI skills, inclusion, and ethical adoption. While they advance technical capacity-building, few explicitly address prompt literacy, the essential skill of interacting responsibly and effectively with AI systems. The AI Dialogue can build on these efforts by connecting stakeholders to share best practices, training resources, and frameworks for responsible prompting. By recognizing prompt literacy as a key governance layer, it complements existing technical and policy programs, ensuring AI is used safely, equitably, and effectively. The Dialogue's added value lies in its ability to coordinate international collaboration, elevate awareness of responsible prompting, and embed it into broader AI governance strategies. This includes joint initiatives to train underrepresented groups, standardize safe prompting practices, and research human-AI interaction risks. By linking global initiatives with a focus on how humans engage with AI, the Dialogue can reduce misuse, strengthen trust, and empower users worldwide.

Different stakeholders can advance prompt literacy through complementary roles. Governments can provide policy support and resources; civil society organizations like AI SISTERS can lead training and mentorship for underserved communities; private sector can offer tools and expertise; and academia can develop curricula and evidence-based guidelines. The AI Dialogue could use a hybrid format: plenary sessions for policy and best practices, focused workshops to co-create prompt literacy frameworks, interactive labs to test tools, and mentorship circles connecting experts with underserved communities. Emphasizing collaboration and inclusivity, each session should produce tangible outputs like open-access guidelines, toolkits, and training resources. By focusing on how humans interact with AI, the Dialogue can bridge governance gaps, empower underrepresented groups, and ensure AI is used safely, effectively, and equitably across all communities.

Voices from women in tech, youth, and underserved communities, especially in the Global South, remain underrepresented in global AI governance discussions. These groups often lack access to AI education, tools, and guidance on responsible use, including prompt literacy, which limits their ability to safely and effectively interact with AI systems. To include them, initiatives like AI SISTERS can be scaled globally, offering training, mentorship, and awareness programs that empower underrepresented communities with prompt literacy skills. Global forums and dialogues can actively invite participants from these groups, provide funding or virtual access, and create dedicated tracks where their experiences, challenges, and ideas inform governance frameworks. By centering these voices and equipping them with skills to engage responsibly with AI, global discussions can become more inclusive, equitable, and effective in shaping safe and human-centered AI.

To foster meaningful engagement during the AI Dialogue, innovative formats should prioritize interaction, collaboration, and inclusivity, particularly around prompt literacy. 1. Interactive Workshops: Hands-on sessions where participants learn and practice responsible prompting, co-create guidelines, and test AI tools in real scenarios. This helps bridge the gap between knowledge and practical application. 2. Hackathons and Labs: Multi-day collaborative events where teams, including participants from underserved communities, design AI solutions or gamified training modules that emphasize safe and effective human-AI interaction. 3. Mentorship Circles: Small-group discussions connecting experts, practitioners, and novices from diverse backgrounds. These circles can focus on prompt literacy, ethical use of AI, and sharing localized best practices. 4. Simulation Exercises: Realistic scenarios where participants experience the consequences of poor prompting, biased outputs, or Shadow AI, highlighting the importance of human responsibility in AI governance. 5. Open-Access Knowledge Hubs: A digital platform accompanying the Dialogue, offering training materials, prompt literacy toolkits, and discussion forums to extend learning beyond the event and reach underserved communities. 6. Plenary Panels with Interactive Q&A: Instead of one-way presentations, panels can integrate live polling, scenario challenges, and breakout discussions to actively involve all attendees. By combining these formats, the AI Dialogue can move beyond passive participation and empower diverse stakeholders, especially those historically underrepresented to understand, practice, and influence responsible AI use. Focusing on prompt literacy ensures participants not only discuss AI governance but leave equipped to engage safely, effectively, and equitably with AI systems in their own communities.

Several policies, practices, and platforms are helping promote effective AI governance, particularly by addressing human-AI interaction challenges. Capacity-building initiatives, like AI SISTERS, provide training and mentorship to underrepresented groups, equipping them with skills in prompt literacy, ethical AI use, and safe system interaction. These programs bridge knowledge gaps and reduce risks such as misinformation, biased outputs, and Shadow AI in organizations. Organizational policies that mandate guidelines for AI use, data handling, and responsible prompting help ensure consistency and accountability. For example, companies can implement prompting best practices, internal audits of AI outputs, and clear escalation procedures when AI generates risky or biased content. Platforms and open-source tools that support responsible AI use such as AI sandbox environments or gamified training modules allow users to practice and test prompts safely, reinforcing both skills and awareness. Collaborative approaches, like multi-stakeholder forums or co-creation workshops, enable governments, academia, civil society, and the private sector to develop shared frameworks for AI governance. By combining these, stakeholders can produce actionable guidelines, training resources, and evaluation mechanisms that are practical, inclusive, and scalable. Overall, these initiatives demonstrate that effective AI governance is not only about regulating algorithms it requires empowering users to interact responsibly with AI systems. Focusing on prompt literacy, structured training, and collaborative frameworks ensures safer, more equitable, and human-centered AI adoption.