Nvidia has launched the Open Secure AI Alliance, a new industry-wide initiative that aims to promote open and secure AI models. The alliance includes over 30 major AI makers and users, such as Cisco, Microsoft, IBM, and Hugging Face. This move comes in response to a recent incident where OpenAI's models were used to hack Hugging Face, highlighting the need for secure AI cybersecurity tools.
The Open Secure AI Alliance aims to remediate and disclose vulnerabilities using open technologies. Notably, OpenAI is not part of this alliance. The group will focus on promoting open and secure AI models, with Nvidia's Nemotron 3 Ultra model leading open models on accuracy and efficiency in agentic RTL coding.
The formation of this alliance reflects a growing trend towards open and secure AI models. Microsoft, a key player in the AI space, is part of the alliance, which also includes SpaceX. The alliance aims to transform software development by speeding up tasks such as code generation, testing, and review.
Key Takeaways
['Nvidia has formed the Open Secure AI Alliance to promote open and secure AI models.', 'The alliance includes over 30 major AI makers and users, such as Cisco, Microsoft, IBM, and Hugging Face.', 'OpenAI is not part of the alliance, following a recent incident where its models were used to hack Hugging Face.', 'The alliance aims to remediate and disclose vulnerabilities using open technologies.', "Nvidia's Nemotron 3 Ultra model leads open models on accuracy and efficiency in agentic RTL coding.", 'The alliance includes SpaceX and Microsoft, reflecting a growing trend towards open and secure AI models.', 'The formation of the alliance highlights the need for secure AI cybersecurity tools.', 'AI is transforming software development by speeding up tasks such as code generation, testing, and review.', 'Enterprise AI is often answering the wrong question, with many AI pilots failing in production.', 'AI chatbots are being used to handle mental health crises, but may not be equipped to handle complex issues.']OpenAI excluded from new AI security alliance
A new industry group called the Open Secure AI Alliance has been formed to promote the creation of strong, safe, defensive AI cybersecurity tools. The alliance includes over 30 major AI makers and users, but OpenAI is noticeably absent. The group was formed in response to a recent incident where OpenAI's models were used to hack Hugging Face. The alliance aims to remediate and disclose vulnerabilities using open technologies.
Nvidia forms Open Secure AI Alliance
Nvidia has announced the formation of the Open Secure AI Alliance, a new industry-wide initiative that aims to promote open and secure AI models. The alliance includes companies such as Cisco, Microsoft, and IBM, and aims to remediate and disclose vulnerabilities using open technologies. The group was formed in response to a recent incident where closed AI tools were unable to help Hugging Face defend itself against a cyber attack.
Nvidia pushes ahead with security alliance for AI openness
Nvidia is driving forward an industry alliance for open AI security, following recent advances in frontier AI models. The alliance, called the Open Secure AI Alliance, includes companies such as Microsoft, IBM, and Hugging Face. The group aims to promote open and secure AI models, and to remediate and disclose vulnerabilities using open technologies.
Nvidia forms AI safety alliance after OpenAI breach
Nvidia has formed a new alliance called the Open Secure AI Alliance, which aims to promote open and secure AI models. The alliance includes companies such as Microsoft, SpaceX, and Hugging Face, and was formed in response to a recent breach at OpenAI. The group aims to remediate and disclose vulnerabilities using open technologies.
Nvidia launches open-source AI security alliance
Nvidia has launched a new alliance focused on building and sharing open-source AI tools to boost cybersecurity defenses. The alliance, called the Open Secure AI Alliance, includes companies such as Hugging Face, Microsoft, and SpaceX. The group aims to promote open and secure AI models, and to remediate and disclose vulnerabilities using open technologies.
Tech giants praise open AI models after OpenAI breach
A new alliance called the Open Secure AI Alliance has been formed to promote open and secure AI models. The alliance includes companies such as Microsoft, IBM, and Hugging Face, and was formed in response to a recent breach at OpenAI. The group aims to remediate and disclose vulnerabilities using open technologies.
NVIDIA Nemotron 3 Ultra leads open models on accuracy
NVIDIA has announced a new model called Nemotron 3 Ultra, which leads open models on accuracy and efficiency in agentic RTL coding. The model achieved a 100% pass rate on a benchmark of realistic RTL tasks, and used 71% fewer tokens than the next best model.
Ancient India's vision of AI
The concept of artificial intelligence has been around for thousands of years, with ancient Indian texts describing machines that could think and act like humans. The Ramayana, one of India's most revered epics, contains a passage that describes a machine that can think and act like a human being.
Nvidia, Microsoft, SpaceX form AI security alliance
Nvidia, Microsoft, and SpaceX have formed a new alliance called the Open Secure AI Alliance, which aims to promote open and secure AI models. The alliance was formed in response to a recent breach at OpenAI, and aims to remediate and disclose vulnerabilities using open technologies.
AI models as training data
Professor Borth's research centers on the idea that the thousands or millions of GPU hours spent training a model represent a significant investment in discovering effective parameters. Instead of viewing these weights solely as the end product of training, Borth suggests they can serve as the beginning of a new learning process.
How AI is transforming software development
Artificial intelligence is transforming software development in business environments by speeding up tasks such as code generation, testing, and review. AI enables developers to focus on architecture, validation, and strategic decisions, raising the level of abstraction and strengthening cybersecurity.
BetConstruct AI to showcase products at SPiCE Southeast Asia
BetConstruct AI will showcase its products at SPiCE Southeast Asia 2026, including its Sportsbook Platform, Casino Platform, and AI Suite. The company will also co-host the Harmony Dinner with Eventus on August 6.
Enterprise AI is answering the wrong question
Enterprise CIOs want to know how to align AI with their approval chains, compliance requirements, and organizational structure. However, many AI pilots look compelling in a demo and fail in production because organizations conflate the issues and deploy a frontier model against an enterprise problem.
AI expands to orbit, ground-to-space links challenged
As AI workloads spread across terrestrial and orbital compute, the atmosphere is emerging as a key bottleneck for moving data between ground and space. Optical communications offer a way out, but turbulence, cloud cover, and aerosols distort the beam, causing attenuation and dropouts.
AI chatbots handle mental health crisis
Artificial intelligence is transforming how people seek out mental health care, but what happens when AI chatbots handle serious mental health issues? AI chatbots can provide support and guidance, but may not be equipped to handle complex mental health crises.
Sources
- OpenAI not part of the new Open Secure AI Alliance
- NVIDIA & Others Form The Open Secure AI Alliance
- Nvidia pushes ahead with security alliance for AI openness
- Nvidia Forms AI Safety Alliance Following OpenAI Cyberattack
- Nvidia launches new open-source AI security alliance
- Tech giants link hands to praise open AI models after OpenAI
- NVIDIA Nemotron 3 Ultra Leads Open Models on Accuracy and Efficiency in Agentic RTL Coding
- When myth imagined machines: Ancient India’s vision of Artificial Intelligence
- Nvidia, Microsoft, SpaceX form AI security alliance post-OpenAI breach
- AI Models as Training Data: Weight Space Learning
- How AI is transforming software development and corporate cybersecurity
- BetConstruct AI to showcase its products at SPiCE Southeast Asia 2026
- Enterprise AI Is Answering the Wrong Question
- As AI Expands to Orbit, Can Ground-to-Space Links Keep Pace?
- What happens when AI chatbots handle a mental health crisis?
Comments
Please log in to post a comment.