Snap Inc. & HackerOne
Pioneering AI Red Teaming: a decade-long partnership driving innovation in AI safety and security through human ingenuity and cutting-edge methodologies.
Leading in AI Red Teaming
Snap has been an early adopter of AI red teaming, working with HackerOne to test and refine strict safeguards for generative AI technologies. Together, we've developed innovative methodologies to surface previously unknown vulnerabilities in AI systems, contributing to a safer, more ethical digital landscape.
AI Safety Red Teaming
Focuses on preventing the generation of harmful content, such as offensive language or instructions for dangerous activities.
Focuses on preventing the generation of harmful content, such as offensive language or instructions for dangerous activities.
AI Security Red Teaming
Ensures bad actors can't exploit AI systems to compromise confidentiality, integrity, or availability.
Ensures bad actors can't exploit AI systems to compromise confidentiality, integrity, or availability.
How Snap protects its innovative products
Innovative CTF-Style Exercises
These exercises incentivized researchers to uncover vulnerabilities in Snap's generative AI products, such as the Lens and My AI Text2Image features, targeting harmful imagery like violence, self-harm, and inappropriate content.
This innovative approach provided valuable insights into AI models' behavior and informed Snap's safety benchmarks, which have become a blueprint for testing harmful content across the tech industry.
Adopting Hai, HackerOne’s AI Security Agent
During a private CTF hackathon, Snap leveraged HackerOne’s AI tool Hai to translate submissions into seven European languages, making it easier to communicate with security researchers worldwide.
Hitting $1M in bounties is a badge of honor. It reflects our commitment to valuing the intelligent security researchers who help keep us safe. Bug bounty pr grams are notoriously difficult to build, but HackerOne’s talented community provides us with the expertise and creativity we need to secure our platform. Our ultimate goal is to make Snap’s bug bounty program a model for others to follow.
AI red teaming allows us to explore the possibilities of what attackers might achieve—not just what’s likely. Working with HackerOne has shown us that human ingenuity often outperforms adversarial datasets or AI-generated attacks.
