Microsoft has announced new security tooling to monitor digital assistants, highlighting concerns that advanced AI models cannot be considered fully reliable.The new 'Guardian Agents' will track assistant activity and introduce requirements for human approval on potentially risky decisions.Microsoft Defender can now identify approximately 60 different types of digital assistants commonly used by employees, while t…https://en.hacks.gr/microsoft-i-techniti-noimosyni-den-einai-apolyta-asfalis-kai-chreiazetai-anthropino-elegcho/#Microsoft #ArtificialIntelligence #DataPrivacy #AI
Microsoft has announced new security tooling to monitor digital assistants, highlighting concerns that advanced AI model...
