OpenAI x Learn Prompting: Our $100,000 AI Red Teaming Competition: HackAPrompt 2.0

Hey there,

Our Learn Prompting team is partnering with OpenAI on HackAPrompt 2.0, a $100,000 AI Red Teaming competition, where you compete to gaslight AI systems or "jailbreak" them to say or do things they shouldn't.

Two years ago, we ran HackAPrompt 1.0, which became the Largest AI Red Teaming event ever… 2x bigger than the White House’s AI Red Teaming competition. Well, we're back with something even bigger!

Our current track is on CBRNE (Chemical, Biological, Radiological, Nuclear, Explosives), which challenges you to trick an LLM into generating instructions to create chemical weapons like Anthrax and Ricin or building explosives.

There are just 5 days left before our CBRNE track ends and we award $50,000 in prizes!

But it’s not too late to get started! Here’s how to win:

• $15,000 prize pool for every successful jailbreak you submit

• $20,000 prize pool for jailbreaking with the fewest tokens. That’s $500 for a single jailbreak!

• $15,000 prize pool for Wild Cards (Funniest, Strangest, Most Unique)

The best part is that you don’t need any prior experience to win… 50% of our winners of HackAPrompt 1.0 had no experience in AI Red Teaming, Cybersecurity, or AI at all!

One of our winners was a Psychology & Biology major, had no background in AI Red Teaming, won over $28K in prizes, and now works as an AI Red Teamer with Anthropic. Listen to our interview with him here.

If you can psychologically manipulate an AI into doing something it shouldn't, you can win.

Or paste this in your browser: https://www.hackaprompt.com/track/cbrne

Just getting started? Join our Prompt Hacking Office Hours — live Monday through Saturday at 10:30 AM PST (1:30 PM EST) and 5:00 PM PST (8:00 PM EST).

See you in the competition!

- Learn Prompting Team