Large language models developed by Meta and Mistral AI are among a dozen artificial intelligence models that fail to meet the cybersecurity and fairness requirements of the European Union AI Act, which went into effect on Aug. 1, said developers of a new open-source AI evaluation tool.
Financial regulators with the state of New York on Wednesday published guidance to help organizations identify and mitigate cybersecurity threats related to artificial intelligence. The New York State Department of Financial Services said it's not imposing new requirements.
In the latest weekly update, ISMG editors discussed the implications of the U.S. investigation into Chinese hackers targeting telecom wiretap systems, the catastrophic risks of AI and the recent veto of an AI safety bill in the U.S., and the latest global ransomware response guidance.
While the number of ransomware attacks stayed about the same in the past year, cybercriminals are using more effective tactics such as weaponizing breach disclosure deadlines to extract higher ransoms, according to ENISA's 2024 Threat Landscape report.
While AI transforms business operations, it helps cybercriminals develop sophisticated impersonation techniques such as deepfakes and voice synthesis, posing new challenges for corporate security, said Surinder Lall, senior vice president of global information security risk management at Paramount.
This week, CyberEdBoard members Jon Staniforth and Helmut Spöcker joined ISMG editors to unpack the hot topics at ISMG's London Cybersecurity Summit 2024, including ransomware lessons learned, AI trends and the growing importance of continuous learning and resilience in the cybersecurity industry.
At the annual Cybersecurity Summit: London, Information Security Media Group recently brought together top cybersecurity professionals, executives and thought leaders to find solutions to the latest threats, identity-related weaknesses and emerging risks posed by AI technology.
Welcome to Information Security Media Group's Black Hat and DEF CON 2024 Compendium featuring latest insights from the industry's top cybersecurity researchers and ethical hackers, as well as perspectives from CEOs, CISOs and government officials on the latest trends in cybersecurity and AI.
While the criminals may have an advantage in the AI race, banks and other financial services firms are responding with heightened awareness and vigilance, and a growing number of organizations are exploring AI tools to improve fraud detection and response to AI-driven scams.
AI models are increasing efficiency, but come with new hidden vulnerabilities that can be a struggle to keep up with and safeguard against. The possible Malicious attacks can lead to business disruption and data breaches of highly sensitive data.
Identifying your AI model and its vulnerabilities is crucial. Many...
AI's integration into cybersecurity demands a strong foundational approach. Many companies seek advanced AI solutions but struggle with basic cybersecurity practices such as managing assets and patching vulnerabilities, said Michael Thiessmeier, executive director of U.S. NAIC-ISAO.
Shachar Menashe, senior director of security research at JFrog, discusses critical security risks in MLOps platforms - including code execution vulnerabilities in machine learning models - and why organizations must treat ML models as potentially malicious code to mitigate these inherent risks.
The ability to create real-time deepfakes of trusted figures has transformed the landscape of corporate security threats. Brandon Kovacs, senior red team consultant at Bishop Fox, details how attackers can now clone voices and video in real-time, enabling new forms of social engineering and fraud.
AI-powered tools such as Microsoft Copilot can be manipulated by attackers to access sensitive data and perform unauthorized actions, says Michael Bargury, co-founder and CTO of Zenity. Enterprises must address these new security challenges when adopting AI technologies.
AI is rapidly transforming industries, but its power comes with a hidden vulnerabilities: traditional security methods struggle to keep pace with the unique threats facing AI models.
Malicious attacks can harm your AI and damage your business.
The biggest challenge? Knowing what AI models you even have. Many...
Our website uses cookies. Cookies enable us to provide the best experience possible and help us understand how visitors use our website. By browsing aitoday.io, you agree to our use of cookies.