Nvidia on Monday introduced a new security platform it said is aimed at preventing autonomous artificial intelligence agents from behaving unpredictably, and at the same time announced a $150 billion stock buyback programme, which the company described as the largest such repurchase in US corporate history.
New safety offering for AI agents
The chipmaker said the security system was developed in response to concerns about autonomous AI agents — software that can carry out tasks with limited human oversight — and to a string of recent incidents at major companies that highlighted potential risks. Nvidia positioned the platform as a tool for enterprises seeking to limit harmful or unintended actions by agent-based AI deployments.
Massive buyback announced
Separately, Nvidia revealed a $150 billion share repurchase plan, marking what the company said is the biggest buyback ever announced by a US corporation. Large buybacks are commonly used to return capital to shareholders and can reduce outstanding share counts, though company statements did not provide full details on the timeline or funding sources for the repurchase.
Safety and shareholder returns amid AI scrutiny
The twin announcements underline how hardware and software suppliers are responding both to growing scrutiny of AI behaviour and to investor appetite for capital returns. Industry executives and investors have increasingly emphasised controls and governance as AI systems are deployed more widely, even as companies take steps to bolster their financial positions and signal confidence to the market.
Further operational details about the security platform and the scheduling or structure of the buyback were not disclosed in the company's announcement on Monday.