Meet with Alice
About the Event
The Trust & Safety community gets its prime gathering of the year - TrustCon 2026, July 20–22 in San Francisco.
The global gathering for the people quietly keeping platforms sane: policy, moderation, integrity, responsible AI and more. Very much our kind of room.
Swing by the morning coffee break (yes, the good cups are on us)!
.jpg)
People for Profit: Scaling Online Trafficking Signals
Traffickers are increasingly using AI to scale recruitment, promotion, and exploitation across multiple forms of trafficking, including sex trafficking, labor trafficking, criminal exploitation, and human smuggling. Using the U.S.-Mexico border as a case study, this session explores how AI-generated text, images, and audio are reshaping trafficking operations and reveals the behavioral signals that remain consistent across abuse types. Through real-world threat intelligence case studies, attendees will learn practical, signal-based frameworks for identifying and disrupting trafficking-related content at scale.
The Step-by-Step Evolution of AI Use in Child Exploitation: from Cheapfakes to Multimodal Abuse
What happens when genuine offender communities adopt generative AI? Drawing on more than 2,000 AI-related discussions from dark web communities, this session provides a rare evidence-based look at how AI is reshaping child exploitation. Attendees will learn how offenders are testing AI models, bypassing safeguards, and using synthetic media to enable grooming, sextortion, and trafficking. The session concludes with practical insights into the evolving tactics, behavioral signals, and detection strategies trust and safety teams can use to identify and disrupt AI-enabled exploitation.
.jpg)
Filling the Vacuum: How Russian Disinformation Actors Are Benefitting From Information Vacuums Through Large Language Models
Large language models are rapidly becoming the go-to source of information for millions of people worldwide, but they have a blind spot: When there is an information vacuum, the LLM will quote whatever scarce sources are available. We see Russian state-aligned disinformation actors benefiting from this weakness. Pro-Russia misinformation narratives promoted into low-competition information spaces are quoted by the world’s most popular LLMs, chatbots and AI Search summaries. I’ll present original benchmarking research across the world's top LLMs and chatbots, testing each against a set of verified Russian disinformation claims, narratives directly traceable to Kremlin-aligned outlets and known influence operations. The findings show that across multiple of the LLMs, these claims are being reproduced, paraphrased, or cited. Through the LLM's citation, the claim also gains an air of credibility through being delivered by an AI system: what remains looks like neutral, reliable information.
Meet the Alice team
