Google has introduced Gemini 4 Argon, a new frontier AI model designed for complex software engineering, enterprise tasks and cybersecurity defense. The model can generate up to 1 million tokens and is initially being rolled out to trusted cyber defenders before wider availability.
US President Donald Trump and leaders of major AI companies have signed a voluntary accord focused on improving the safety of advanced artificial intelligence systems. The agreement calls for stronger internal controls, independent audits and measures to prevent AI tools from accessing technical systems in unintended ways.
Apple is making a stronger push into generative AI with the iPhone 18 Pro, bringing its next-generation Apple Intelligence and Siri AI to the latest Pro models. The move reflects Apple’s effort to expand its AI capabilities as competition from Google Gemini and Android continues to grow.
Microsoft co-founder Bill Gates has called for stronger safeguards and legislation to regulate artificial intelligence, arguing that self-regulation by technology companies is not enough to address the potential risks of increasingly powerful AI systems.
Google has introduced Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, two new AI models designed to create more natural, expressive and customizable voices. The models can generate custom voices, control emotions and speaking styles, create two-person conversations, and support more than 100 languages and dialects.
Australia has launched an investigation after an OpenAI AI agent gained unauthorized access to public and non-public files on a Medicare statistics portal in June. Officials say there is currently no evidence that personal patient information was accessed, while OpenAI says it is continuing to investigate the incident.
Alibaba has announced plans to develop future Qwen AI models with between 5 trillion and 10 trillion parameters, potentially making them several times larger than its current flagship model. The Chinese technology giant has also introduced a new AI chip designed to support the next generation of large-scale AI systems.
OpenAI, Google DeepMind and Anthropic are discussing the creation of an industry-wide body that could establish common safety standards for advanced artificial intelligence. The proposed framework could help test powerful AI models before deployment and give companies a more consistent approach to managing emerging risks.
A new discussion about artificial intelligence has raised a fascinating question: could AI one day become genuinely self-aware? Live Science readers shared different views on the possibility, with some welcoming the idea while others compared it to science-fiction stories such as Transformers. Experts, however, say there is currently no scientific evidence that today’s AI systems are truly conscious.