← back
Prompt Engineering and AI Red Teaming — Sander Schulhoff, HackAPrompt/LearnPrompting
Takeaway
Prompt injection is a fundamentally unsolved AI-security problem and red-teaming competitions like HackAPrompt are the empirical baseline for measuring it.
Summary
- Sander Schulhoff (LearnPrompting/HackAPrompt) argues prompt engineering isn't dead—it gates secure deployment of LLMs and agents.
- Background spans deception research on Diplomacy, MineRL Minecraft agents, and authoring the first internet prompt-engineering guide (cited by OpenAI, Google, NIST, USG).
- HackAPrompt ran the first AI red-teaming competition; the resulting 600k-prompt dataset is used by every major AI lab to benchmark robustness.
- Frames AI security as an essentially unsolvable problem distinct from classical cybersecurity, with prompt injection as the dominant new attack class.
prompt-injectionred-teamingsecurity
Original description
Learn from the creator of Learn Prompting, the internet's 1st Prompt Engineering guide (released 2 months before ChatGPT), and HackAPrompt, the World's 1st AI Red Teaming competition. My talk will cover topics ranging from the history of prompt engineering to the most advanced research-backed prompt engineering techniques. I will also discuss the origins of prompt injection and AI red teaming, as well as the current state of industry and the need for agentic red teaming. https://www.hackaprompt.com About Sander Schulhoff I'm Sander Schulhoff, the founder and CEO of HackAPrompt and Learn Prompting. I created the first Prompt Engineering guide on the internet, two months before ChatGPT was released, which has taught 3 million people how to prompt ChatGPT. I also ran, in collaboration with OpenAI, the first AI Red Teaming Hackathon (an event that nearly doubled a similar one by the White House). Today, HackAPrompt partners with the Frontier AI labs to produce research that makes their models more secure. My background is primarily in NLP and deep reinforcement learning. I recently led the team behind The Prompt Report, the most comprehensive study of prompt engineering ever done. Our 76-page survey, co-authored with OpenAI, Microsoft, Google, Princeton, Stanford, and other leading institutions, analyzed 1,500+ academic papers and covered 200+ prompting techniques. Recorded at the AI Engineer World's Fair in San Francisco. Stay up to date on our upcoming events and content by joining our newsletter here: https://www.ai.engineer/newsletter