Neural networks block website copying for security reasons

The largest language models, Claude and ChatGPT, have implemented new restrictions that block users' attempts to copy the code of websites and mobile applications. AI assistants reject such requests, citing internal cybersecurity policies.

The catalyst for this tightening of policy was an incident involving a researcher who successfully removed ads and tracking features from Samsung TVs in just 11 hours with a budget of $200. The device was connected to hundreds of partners, causing a stir in the security community.

Now, when attempting to recreate website interfaces or code, artificial intelligence may interrupt the dialogue. Requests are classified as Cyber Abuse, making automated parsing and reverse engineering via chatbots impossible.