Nvidia-led AI security group proposes incident-reporting standards

The Open Secure AI Alliance (OSAA), formed in late July 2026 under Nvidia's leadership, has launched a working group called the Shared AI Findings Exchange (SAFE) and begun presenting proposals for open comment.

Developed during the Black Hat conference the week of August 4, 2026, the proposals are being managed by the Linux Foundation.

The guidelines focus on confidentially reporting AI cybersecurity incidents, alerting affected parties, and conducting blame-free analysis to foster collective learning.

More than 120 companies have joined OSAA, including Adobe, BlackRock, Cisco, Intel, Microsoft, and Visa. Anthropic, OpenAI, and Google are notably absent.

Both OpenAI and Google signed the original open letter that inspired the group, published in the week of July 27, 2026. The letter urged the White House to support open source AI efforts. Anthropic did not sign.

Members are also contributing open source technology: Nvidia's open models and vulnerability scanner Garak, Okta's agent identity technology, Red Hat's agent governance work, and Amazon's Strands Agents tool and Cedar authorization language.

The rapid action follows industry concern over the Trump Administration's potential ban on Chinese open weight models, which prompted the open letter.

“Openness may be one of the most important paths to AI safety and security,” the group wrote in their letter.