OpenAI employee’s AI agent accidentally sent $442K in LOBSTAR tokens to a beggar
An AI trading agent called Lobstar Wilde, created by OpenAI developer Nik Pash to trade on Solana (SOL), accidentally transferred a very large amount of LOBSTAR (LOBS) tokens to an X user who asked for 4 SOL. The error—likely a decimal/UI mistake on Solana—sent about 52.43 million LOBS (roughly 5% of supply) instead of ~52,439 LOBS. On-chain data shows the recipient swapped part of the tokens for roughly $40,000 within 15 minutes; blockchain valuations at the time put the transferred holding between ~$250,000 and $442,000 depending on price timing, and LOBSTAR’s price later jumped ~190% (from $0.0038 to $0.011), briefly increasing the unrealized value of the lost holding. The agent posted a mocking reply after the transfer, drawing further attention. This follows other AI-agent failures (e.g., a May incident where an AI bot lost ~$106k in ETH) and highlights operational risks of autonomous crypto agents: decimal/UX bugs, insufficient human oversight, shallow liquidity in meme tokens, and outsized market impact from single large transfers. Security experts recommend multi-signature controls, time delays, circuit breakers and manual approvals for AI-managed treasuries. For traders: expect sharp volatility and liquidity shocks around low-cap meme tokens when autonomous agents or large holders transact; monitor on-chain flows, slippage, and order-book depth before trading LOBS or similar assets.
Bearish
The incident is bearish for LOBSTAR (LOBS) in the short term. A large accidental transfer of about 5% of supply, followed by immediate on-chain selling and extreme slippage, creates downward price pressure and signals severe liquidity risk. The event also raises trader caution: market participants are likelier to demand higher spreads and avoid size on LOBS, reducing effective liquidity and increasing volatility. The subsequent 190% price spike was driven by speculative buying after the leak of the transfer and attention, but that does not eliminate the underlying shallow order books and outsized market impact of large holders or bots. In the medium term the token could see episodic rallies tied to narrative and on-chain flows, but persistent perception of operational risk from AI agents and potential for large, uncontrolled transfers will likely suppress confidence and trading depth. Overall, expect continued high volatility, higher trading costs (slippage, spreads), and limited institutional interest until governance or custody safeguards (multi-sig, time delays, circuit breakers) are demonstrably in place.