Get paid $30,000 to hack Google’s AI — legally

Move over, Silicon Valley code jockeys — there’s a new side hustle in town. Google just launched an AI Vulnerability Reward Program, dangling up to $30,000 for anyone smart (or devious) enough to find bugs in its artificial intelligence systems. Think of it as a digital treasure hunt, except the “treasure” is Google admitting its AI isn’t as bulletproof as it looks.

What Google’s AI Vulnerability Reward Program is about

Before you start typing prompt injections into Gemini or trying to make Bard order you pizza, a warning: this bounty isn’t for catching your chatbot saying something stupid. Google doesn’t care if its AI hallucinates a fake Wikipedia page or calls you “bro.” The company wants real vulnerabilities — the kind that could let an attacker siphon private data, hijack user accounts, or get an AI to perform unauthorised actions. Basically, if you can make the machine do something bad, not just say something dumb, you’re in the money.

The payout scale reads like a gamer’s leaderboard: $20,000 for high-impact finds in Google’s crown jewels (Search, Gmail, Gemini), up to $30,000 for originality and precision. Bugs in side projects like NotebookLM earn less. But bragging rights? Priceless.

Critics say thirty grand is peanuts for exploits that could, in theory, compromise millions. Still, it’s a step toward the wild west of AI security — where companies finally admit their models can be hacked, gamed, or socially engineered. Google claims it’s already paid $430,000 for AI-related flaws under its broader bug bounty, but this is the first time it’s ring-fencing cash specifically for AI.

So yes, you can now get paid to outsmart the world’s biggest tech giant — as long as you play by the rules. Find the cracks, file the report, cash the check. Silicon Valley’s most advanced minds built these AIs; now they’re paying you to break them.

For more updates, join/follow our WhatsApp, Telegram and YouTube channels.