Tag: AI safety

Here you will find all the articles with the tag: AI safety.

New tools in Microsoft Azure AI Studio keep LLMs safe

New tools in Microsoft Azure AI Studio keep LLMs safe

A set of newly unveiled tools should make models in Azure AI Studio more accurate and secure. Microsoft is addressing a number of important AI vulnerabilities with the update. The update brings five safety tools to Microsoft Azure AI Studio. First, Prompt Shields should enable organizations to p... Read more

date19 days ago
How AI can be trained to deceive, even after safety training

How AI can be trained to deceive, even after safety training

LLMs can engage in intentional malicious behaviour, even when current methods to promote their safety are employed. Such behaviour can be deeply persistent, Anthropic research shows. Anthropi'sc researchers have presented these findings in their paper "Sleeper Agents: Training Deceptive LLMs tha... Read more

date3 months ago
Cloud Security Alliance expands into AI security

Cloud Security Alliance expands into AI security

The Cloud Security Alliance (CSA) is now expanding its work to encompass AI security. The CSA recently launched the AI Safety Initiative for this purpose. Many major (cloud) tech companies are collaborating on standards and best practices for cloud security within the CSA. The group argues that ... Read more

date4 months ago