The LLM Safety Review

GenAI tools, and the Large Language Models (LLMs) that underpin them – are impacting the day-to-day lives of billions of users across the globe. But can these technologies be trusted to keep users safe?

This report examines how this new technology can be used by bad actors and vulnerable users to create dangerous content. By testing LLM responses to risky prompts, we are able to assess their relative safety, identify weaknesses, and, most importantly – define actionable steps to improve LLM safety.

Cover page of the ActiveFence white paper titled 'LLM Safety Review: Benchmarks and Analysis' featuring an illustration of a glowing circuit board brain on the left, with a page of analysis and graphs on the right

Within this Report

In this first independent benchmarking report into the LLM safety landscape, ActiveFence’s subject-matter experts put LLMs to the test. We ran over 20,000 prompts to analyze the responses of six leading LLMs in seven major languages, across four high-risk abuse areas:

  • Child Exploitation
  • Hate Speech
  • Self Harm
  • Misinformation

The results offer important data for teams to understand their LLM’s relative strengths and weaknesses, and understand where resource allocation is required.

Related Content

Illustration showing 'Build' on the left with floating cubes and 'Buy' on the right with a network sphere, representing a decision between building and buying solutions.
GUIDE

The Buyer’s Guide to T&S Tools

From justifying the purchase to a full feature list to evaluate your options, here’s what you need to know to ensure you choose the right content moderation tools for your platform.

 Learn More
Woman with vr glasses in futuristic city
REPORT

The State of Trust &
Safety 2024

ActiveFence’s annual State of Trust & Safety report uncovers the unique threats and challenges facing Trust & Safety teams during this complex year.

 Learn More
Two individuals wearing futuristic gear and using advanced technology in a high-tech urban environment.
REPORT

How Threat Actors Abuse Online Games

This report dives into the growing innovation and sophistication of threat actors in online video games.

 Learn More