As online content becomes increasingly complex, Trust and Safety teams must evolve their moderation strategies to keep pace. Emerging technologies offer new opportunities to enhance detection and response capabilities, maintaining safety at scale.

• Harnessing multi-modal AI and LLMs to analyze text, images, and video for more accurate content moderation.
• Implementing detection mechanisms to identify and respond to AI-generated content, including deepfakes and synthetic media.
• Advancing real-time moderation of visual content to combat harmful imagery and policy violations instantly.
• Preparing for the next wave of challenges in online trust and safety, from evolving AI threats to regulatory shifts

This closing panel discussion will explore the critical challenges and solutions in scaling Trust & Safety operations amidst rapid platform growth and evolving responsibilities. Attendees will gain insights into balancing policies, moderation, and enforcement at scale.

• Examining strategies for effective resource allocation and efficiently scaling T&S teams and technologies to match user growth.
• Balancing human moderation and AI-powered solutions to handle increasing content moderation requirements.
• Embedding T&S considerations across all business units, ensuring that safety and integrity are prioritised throughout the organization’s growth.
• Evolving T&S policies to accommodate growth while maintaining consistency and fairness across diverse user bases.
• Evaluating the effectiveness of T&S efforts at scale, including key performance indicators and metrics that reflect both quantitative growth and qualitative improvements in user safety

For years, hashing technology has helped platforms detect and remove known child sexual abuse material (CSAM), but the challenge of identifying new, unknown CSAM has persisted—until now. Recent advancements in AI are transforming child safety efforts, enabling platforms to proactively detect previously unreported CSAM at scale. This panel brings together experts from Thorn, Internet Watch Foundation and Hive to explore how AI classifiers are reshaping the fight against online child exploitation. 

• How AI models are trained to detect unknown CSAM while minimizing false positives and negatives
• The role of human oversight in AI-driven content moderation
• Strategies for increasing industry adoption and ensuring AI tools evolve responsibly
• The emerging challenge of AI-generated CSAM and what platforms can do to prepare