Microsoft has tricked several gen-AI models into providing forbidden information using a jailbreak technique named Skeleton Key.
The post Microsoft Details ‘Skeleton Key’ AI Jailbreak Technique appeared first on SecurityWeek.
Tech Leaders Gather This Week for AI Risk Summit + CISO Forum at the Ritz-Carlton, Half Moon Bay
SecurityWeek’s AI Risk Summit + CISO Forum brings together business and government stakeholders to provide meaningful guidance on risk management and cybersecurity in the age of artificial intelligence.
The post Tech Leaders Gather This Week for AI Risk Summit + CISO Forum at the Ritz-Carlton, Half Moon Bay appeared first on SecurityWeek.
OpenAI Co-Founder Sutskever Sets up New AI Company Devoted to ‘Safe Superintelligence’
Ilya Sutskever’s new company is focused on safely developing “superintelligence” – a reference to AI systems that are smarter than humans.
The post OpenAI Co-Founder Sutskever Sets up New AI Company Devoted to ‘Safe Superintelligence’ appeared first on SecurityWeek.
AI Weights: Securing the Heart and Soft Underbelly of Artificial Intelligence
AI model weights govern outputs from the system, but altered or ‘poisoned’, they can make the output erroneous and, in extremis, useless and dangerous.
The post AI Weights: Securing the Heart and Soft Underbelly of Artificial Intelligence appeared first on SecurityWeek.
Tech Leaders to Gather for AI Risk Summit at the Ritz-Carlton, Half Moon Bay June 25-26, 2024
SecurityWeek’s AI Risk Summit + CISO Forum bring together business and government stakeholders to provide meaningful guidance on risk management and cybersecurity in the age of artificial intelligence.
The post Tech Leaders to Gather for AI Risk Summit at the Ritz-Carlton, Half Moon Bay June 25-26, 2024 appeared first on SecurityWeek.
CISA Conducts First AI Cyber Incident Response Exercise
The US cybersecurity agency CISA has conducted a tabletop exercise with the private sector focused on AI cyber incident response.
The post CISA Conducts First AI Cyber Incident Response Exercise appeared first on SecurityWeek.
Aim Security Raises $18M to Secure Customers’ Implementation of AI Apps
Aim Security has raised a total of $28 million to date and is on a mission to help companies to implement AI products with confidence.
The post Aim Security Raises $18M to Secure Customers’ Implementation of AI Apps appeared first on SecurityWeek.
OpenAI Appoints Former NSA Director Paul Nakasone to Board of Directors
Retired U.S. Army General Paul M. Nakasone brings cybersecurity experience to OpenAI’s Board of Directors and Safety and Security Committee.
The post OpenAI Appoints Former NSA Director Paul Nakasone to Board of Directors appeared first on SecurityWeek.
Microsoft Delaying Recall Feature to Improve Security
Microsoft is not rolling out Recall with Copilot+ PCs as it’s seeking additional feedback and working on improving security.
The post Microsoft Delaying Recall Feature to Improve Security appeared first on SecurityWeek.
Easily Exploitable Critical Vulnerabilities Found in Open Source AI/ML Tools
Protect AI warns of a dozen critical vulnerabilities in open source AI/ML tools reported via its bug bounty program.
The post Easily Exploitable Critical Vulnerabilities Found in Open Source AI/ML Tools appeared first on SecurityWeek.
