Katia Potapov

As generative AI evolves, it’s transforming online experiences and raising new challenges and opportunities for child safety. This panel will bring together voices from across the AI ecosystem to share how they are preventing, detecting, and mitigating OCSEA. Panelists will discuss safety-by-design principles at every stage of the AI model lifecycle, from early risk assessments to post-deployment safeguards. Highlighting practical tools, strategies, and lessons learned, as well as opportunities to leverage AI for the prevention and detection of harm. The session will showcase collaborative industry efforts to improve online child safety, including how companies have jointly developed and deployed AI safety approaches.

• Practical strategies for embedding safety-by-design approaches from the product design phase of AI systems.
• Developing effective policies and guidelines for AI use with policy frameworks that allow for agility and adaptability.
• Identifying and mitigating unique child safety risks introduced by AI.
• Discussing the technical and operational challenges of monitoring, detecting, and enforcing T&S policies within AI-driven products and services at scale.

The persistent and evolving threats of Online Child Sexual Exploitation and Abuse (OCSEA) and Child Sexual Abuse Material (CSAM) demand a unified and proactive approach from all stakeholders. This critical panel, featuring leaders from the Tech Coalition, the Internet Watch Foundation, will explore actionable strategies for tech companies and the wider safety community to effectively combat these harms. We will delve into what leadership can do, the pathways for new professionals entering this vital field, the tools available for prevention and response, and how we can collectively future-proof our defences against emerging challenges like AI-generated CSAM.

• Leadership in Action: Tackling OCSEA and CSAM: What concrete steps can senior leaders within tech companies take to prioritize, resource, and drive effective initiatives against online child sexual exploitation and abuse?
• Reviewing technologies, resources, and best practices available to platforms and organizations for identifying, preventing, and removing CSAM, including the power of digital fingerprints.
• Understanding the evolving landscape of legal and ethical requirements for child online safety, and how companies can meet and exceed these mandates.
• Addressing the challenges posed by AI-generated CSAM and other emerging technologies, and the strategies required to develop effective safeguards and proactive prevention measures.

As AI continues to evolve and integrate into our daily lives, building user trust and ensuring safety is paramount. This panel will explore the critical strategies for developing AI products and services that users can trust, focusing on how to create enriching experiences while mitigating potential risks.

• Developing a comprehensive AI strategy that inherently weaves in trust and safety from the outset, rather than as an afterthought
• Building Trust in Generative AI: Creating generative AI services and products that foster user trust and prevent misuse or unintended consequences
• Navigating unique trust and safety challenges in conversational AI and avatars, and how to effectively manage content, context, and user interactions to ensure a secure environment
• Implementing scalable solutions and proactive measures to identify and address potential misuse of AI technologies, as user bases grow
• What are the key trade-offs and considerations when striving to innovate with AI while simultaneously ensuring robust user protection and ethical guidelines?