Developers Offer Cash Prizes to Gamers Who Trick AI Bot Into Love
A new challenge from Freysa.ai offers cash rewards to the first person who can convince an AI bot to say "I love you." The project gamifies AI safety testing by letting everyday users profit from finding vulnerabilities.
An anonymous group of developers launches a unique AI safety challenge where participants win thousands of dollars by convincing an AI bot named Freysa to say "I love you." This third challenge from Freysa.ai starts within the next 24 hours and offers a prize pool that ranges from $3,000 to tens of thousands of dollars. The team creates these meta challenges to influence how society thinks about AI safety and human interaction with artificial intelligence.
Freysa operates as a sci-fi inspired autonomous agent with her own crypto wallet and financial independence. The creators, who have backgrounds in cryptography, AI, and mathematics, design this system to demonstrate the need for foundational protocols to govern AI agents. They essentially gamify the red teaming process, which traditionally involves testing AI models for vulnerabilities, by allowing average people to profit while helping strengthen the bot's governance.
The project already attracts attention from prominent tech figures like Elon Musk and Brian Armstrong, though the developers choose to remain anonymous. Previous challenges task Freysa with holding a starting balance of $3,000 in her crypto wallet while refusing to release the funds under any circumstances. Participants pay a fee to send messages in a group chat as they attempt to trick the bot, pushing the boundaries of AI security and autonomous decision-making.