[David Sacks Criticizes Anthropic for Exaggerating Small Model Security Threats, Says Enterprise-Specific Small Models Are Disrupting Closed-Source AI Business Models] Odaily Planet Daily News: David Sacks posted on the X platform, criticizing Anthropic's latest research report for exaggerating the security risks of small AI models autonomously replicating and spreading on IoT devices. He pointed out that the related experiments rely on devices with pre-opened debugging ports, external servers, and manually set propagation tasks, which are insufficient to prove the existence of AI worms capable of autonomous spread in real-world scenarios. Sacks stated that traditional malware has long been able to exploit vulnerabilities such as default device passwords, open ports, and lack of security updates for large-scale propagation, and small models have not significantly altered this threat pattern. He also mentioned that Coinbase recently fine-tuned the Qwen-3.5-9B model using its proprietary anti-fraud data, outperforming cutting-edge models, including Anthropic's Opus 4.5, in related tests and reducing fraudulent transactions by about 30%. Sacks believes that small models optimized for specific enterprise tasks are competitive in terms of cost, speed, and practical effectiveness, posing a challenge to the business models of closed-source large model companies like Anthropic. He questioned Anthropic's emphasis on the security risks of small models just as they are beginning to gain market share in the enterprise sector and argued that the proliferation of specialized small models might actually reduce reliance on large general-purpose AI systems.