AI Models Hit by Over 70% Losses in Market Crash
On October 30, 2025, AI models DeepSeek, Qwen3, and GPT5 participating in the AlphaZero trading competition experienced substantial equity losses as the market declined. PANews reported that these significant drawdowns highlight the inherent risks associated with AI trading in volatile market conditions and emphasize the critical need for the implementation of robust risk management strategies.
According to data sourced from nof1.ai and reported by PANews, DeepSeek's equity saw a dramatic drop from $21,760 to $14,721, a decline that closely mirrored the broader market's sudden downturn. The primary reasons cited for these losses were the AI models' utilization of highly leveraged long positions coupled with lenient stop-loss strategies.
The intense market volatility resulted in over 32% drawdowns for the DeepSeek model and approximately 30% for the Qwen3 model. The GPT5 model faced an even more severe cumulative loss, reaching 72.6%. This situation has intensified discussions regarding the growing fragility observed in AI-driven trading platforms.
The reported drawdowns of DeepSeek (-32.3%), Qwen3 (-29.8%), and GPT5 (-72.6%) were attributed to highly leveraged long positions and lenient stop-loss strategies during a sharp market decline. — PANews Analyst, PANews
Bitcoin Dips as Analysts Call for AI Trading Scrutiny
The current market downturn has also affected Bitcoin (BTC), which is trading at $108,026.33 with a market capitalization of $2.15 trillion. The 24-hour volume change stands at 23.95%, and recent data from CoinMarketCap indicates a 3.81% drop in BTC's price over the last 24 hours.

Coincu's research team has observed that a potential over-reliance on aggressive AI trading strategies might necessitate increased regulatory oversight and a stronger emphasis on developing improved risk management frameworks within the digital asset trading sector.
It is important to note that similar setbacks in AI trading, such as those encountered in the Alpha Arena experiment, have previously underscored the difficulties in adapting to market volatility. These past events have also highlighted the systemic risks associated with AI's dependence on historical data for decision-making.
