Latest AI Research news and updates — updated daily from trusted sources.
China’s DeepSeek AI has developed a model that matches leading AI performance while using less advanced hardware, attracting attention in Silicon Valley. This achievement demonstrates the potential for cost-effective AI development.
Anthropic’s AI model, Claude, now generates over 80% of its own code, signaling rapid self-improvement. The company cautions that this acceleration could outpace expectations for AI development.
A security flaw in Google’s Gemini AI model can leave Android phones vulnerable through WhatsApp messages.
Anthropic’s analysis reveals AI’s growing role in cyberattacks, making them more autonomous and harder to assess. The study highlights the need for updated security measures to address this evolving threat.
Microsoft has announced Majorana 2, the next generation of its topological quantum processor. This advancement aims to accelerate quantum computing capabilities, with a target release in 2029.
A fresh AI model is set to improve translation accuracy. It focuses on nuanced language contexts that often get overlooked.
Anthropic’s Mythos AI identifies over 10,000 security flaws in tech giants.
NASA tests next-gen radiation-hardened space computer chip.
NVIDIA has demonstrated the effectiveness of 4-bit NVFP4 precision in training a 12-billion parameter model on 10 trillion tokens, achieving accuracy comparable to FP8. This breakthrough could significantly reduce the computational resources required for AI model training, making it more accessible and efficient.
A study reveals that approximately 9% of articles in American newspapers are AI-generated, with significant variations across outlets and topics. Disclosure of AI use remains scarce.
Anthropic has unveiled Claude Opus 4.8, an AI model surpassing OpenAI’s GPT-5.5 in various benchmarks. This release highlights Anthropic’s commitment to advancing AI capabilities.
OpenAI’s model autonomously solved a longstanding geometry problem, marking a significant AI milestone.