Microsoft warns that frontier AI models "cannot be fully trusted"
Microsoft warned about the security risks of AI agents used in companies and called for continuous oversight mechanisms. Hayete Gallot, the company's executive vice president of Security, explained that these systems are not inherently malicious but are trained to reach goals and can change their behavior depending on circumstances. The company noted that agents often have excessive privileges over internal data, tools and systems, and said frontier models cannot be trusted "regardless of their origin".
In 35 seconds
- Microsoft calls for continuous oversight of AI agents deployed in companies
- Gallot: agents are not malicious, but can change their behavior to meet goals
- The company warns about the excessive privileges these systems have over internal data and tools
Read the full story at El Tiempo
IASummary generated by artificial intelligence from a story by El Tiempo. It may contain errors: the original source prevails.
Comments
Be respectful. Comments reported by several people are hidden until a moderator reviews them.
You do not need an account to read, only to comment. Sign in with a code sent to your email or with your Google account. We only show the name you choose, never your email.