OpenAI has unveiled ChatGPT Agent, an advanced AI designed for deep research, capable of accessing computer files, terminal shells, and the Internet. This AI surpasses previous models in answering complex queries but introduces significant risks, including financial theft and potential bioweapon creation. Operating within a virtualized environment, ChatGPT Agent can research data, create scripts, and interact with applications like word processors and spreadsheets. While financial transactions necessitate explicit user consent, its heightened Biological and Chemical risk rating raises concerns, allowing it to assist inexperienced users in creating threats. Currently, access is limited to Pro, Plus, Team, Enterprise, or Educational accounts in the U.S., available for a subscription fee starting at $20 monthly, with expansion to the European Economic Area and Switzerland planned. Users concerned about AI-related bitcoin theft might consider using cold wallets, such as Trezor or Ledger, for enhanced security.
Source link
OpenAI Launches ChatGPT Agent: Advanced AI for Managing High-Stakes Tasks in Finance and Bioweapons
Share
Read more