The aftermath of the OpenAI cyber attack continues, as NVIDIA, SpaceX, and Microsoft jointly launch an AI security initiative
The aftermath of the cyberattack on OpenAI is still ongoing, as tech giants like NVIDIA, Microsoft, and SpaceX have jointly launched the "Open Security AI Alliance," aimed at using open models to fix vulnerabilities and enhance cybersecurity capabilities. This move responds to the incident where Hugging Face struggled to defend itself due to restrictions on closed models, sparking widespread discussions about AI access limitations and the open-source versus closed-source models
On Monday, NVIDIA, along with several tech giants, launched a new artificial intelligence safety initiative focused on open models. This comes as the aftermath of cyberattacks carried out by uncontrolled OpenAI models continues to unfold.
Last week, it was reported that the target of these attacks—startup Hugging Face—was unable to defend itself using leading-edge models from the U.S. because the guardrail mechanisms of these models could not distinguish between attackers and defenders. As a result, the company turned to a self-hosted, open-weight model that is not subject to the same restrictions.
Open models can be downloaded, modified, and self-hosted, contrasting with closed models (including cutting-edge systems built by Anthropic and OpenAI), which can only be accessed through specific infrastructure.
"The Open Safety AI Alliance will be dedicated to using open technology to fix and disclose vulnerabilities," NVIDIA stated in a press release. "The recent Hugging Face security incident serves as a clear reminder: cyber defenders need open frontier agent systems for self-defense."
In addition to NVIDIA, other members of the alliance include Microsoft, SpaceX, Palantir, and dozens of other tech companies from the U.S. and Europe.
Debate on Restrictions
Currently, there is a growing call for measures to limit access to certain AI models.
Last week, Treasury Secretary Scott Bessen threatened to impose sanctions on companies implementing distillation attacks.
"The U.S. government could indeed implement relevant restrictions," said Chris McGuire, a senior fellow at the Council on Foreign Relations.
He indicated that this could include prohibiting transactions involving relevant models, such as purchasing tokens via API or companies hosting models in the cloud and charging customers for inference fees.
"In Washington, this is not a debate about open source versus closed source; it is a debate about whether to tolerate intellectual property theft," McGuire said. "Any action will target the offending companies, not the open-source ecosystem."
However, concerns about restrictions arise because most of the most powerful open-source models are built by specific companies.
Last week, NVIDIA, Microsoft, Meta, Palantir, and more than 20 other companies jointly issued a letter urging policymakers to avoid imposing "premature restrictions" on open-weight AI models to prevent "stifling competition or driving innovation overseas."
NVIDIA stated that the OpenAI and Hugging Face incidents reveal a "harsh reality": "When defenders cannot inspect, adjust, and run advanced AI on their own infrastructure, their responsiveness is constrained precisely when speed is most needed."
