Cybersecurity
Anthropic Introduces New Protection Against AI Jailbreaks
Anthropic is taking on AI jailbreaks with a powerful new security solution. But can their defense really hold up against the threats of the future?
Cybersecurity
Anthropic is taking on AI jailbreaks with a powerful new security solution. But can their defense really hold up against the threats of the future?
AI News
DeepSeek R1 has quickly become one of the most popular AI models, but now Cisco reveals the chatbot fails to block 100% of malicious requests. At the same time, it has strict limits around politically sensitive topics in China. Is low cost more important than security in AI development?
AI News
Anthropic is making advanced AI models available to the US defense sector in partnership with Palantir and AWS. Is this the next step toward an AI-driven security strategy, and what risks need to be weighed for it to be done safely?
AI News
OpenAI is taking steps toward AI agents that can work autonomously, with launch planned for next year. With the potential to transform everyday tasks and workflows, agents represent a major step forward toward artificial general intelligence.
Cybersecurity
Illegal AI chatbots on the black market are thriving and can generate large sums of money. A study shows these malicious LLMs can be used to create phishing emails and malicious code, posing significant risks to both individuals and companies.
Cybersecurity
OpenAI has implemented several new safety measures to manage the risks around developing and releasing its AI models. This comes through a newly established independent committee, the Safety and Security Committee (SSC), which will oversee safety and security processes. Independent Oversight for Greater Safety The SSC will be an independent
AI Models
AI models can collapse according to a new study, but what does that mean for the future of AI, and how can it be avoided?
Cybersecurity
Criminals are now using AI to improve document forgery and exploit security gaps, increasing the scale and sophistication of their attacks. AI's rapid development challenges security experts and demands new strategies to handle these threats.
Cybersecurity
New studies reveal that AI models like GPT-4 and Meta's Cicero can exhibit deceptive behavior when trained for specific purposes. Although these models don't lie of their own volition, the findings raise concerns about the potential use of AI for mass manipulation.