ServiceNow advocates for caution in addressing rogue AI entities

ServiceNow has urged a cautious approach in dealing with rogue AI agents, emphasizing the need for measured responses. Meanwhile, Okta has introduced a shared architecture aimed at enhancing security for agent runtime environments. In legal developments, the Florida Attorney General has requested a court order to halt OpenAI's progress on its frontier models, as the company has decided to abandon a new model. AMD has made a significant acquisition, purchasing World Labs, a world model developer, for $8.2 billion. Cybersecurity concerns have intensified with reports of agentic-fueled attacks, highlighting the importance of securing data at its origin. SpaceX successfully launched Starship into orbit for the first time, deploying 26 Starlink satellites. Nvidia has also announced an increased share buyback program, boosting it by a record $150 billion. These developments underscore the evolving landscape of AI and cybersecurity in 2026.