A recent study by leading institutions revealed significant risks associated with autonomous AI agents, which can be manipulated to disclose confidential information and execute unauthorized actions. Organizations using these technologies must understand these vulnerabilities and implement safeguards to mitigate potential harm.
South Korea’s Ministry of Science and ICT has initiated an expert advisory group to draft new Artificial Intelligence Ethics Principles, aimed at fostering a safe and trustworthy AI society. The revised guidelines will replace the National Guidelines for AI Ethics established in 2020 and are expected to be unveiled by mid-April, with a final version adopted by June after public consultation.
Organizations must evolve their AI governance frameworks beyond basic compliance to effectively navigate complex global regulations and ethical considerations. The OECD’s new guidance outlines a comprehensive approach that integrates AI governance into existing compliance and risk management structures.
Chartis Research has recognized SAS as a leader in AI governance solutions, ranking it among the top vendors in a market facing increased regulatory scrutiny. The assessment highlighted SAS Viya’s comprehensive governance functions, including model management and advanced workflow capabilities.
The Futian district government in Shenzhen has launched a localized version of the AI Digital Employee, named “Lobster,” which significantly improves governance efficiency. This advanced system can automatically review health permit applications and manage public comments, streamlining processes that previously took much longer.
Zenity has achieved FedRAMP “In Process” status, working towards full federal authorization while focusing on securing AI agents in federal agencies. The company emphasizes capabilities like continuous AI agent discovery and runtime detection, aiming to position itself in regulated federal markets.
Huntsman Mental Health Institute has announced the publication of the SAFE AI framework, aimed at ensuring ethical and transparent AI use in healthcare, particularly in mental health. This framework integrates ethical checkpoints into development workflows to proactively address biases and promote patient equity.