The Future of AI, Automation & Digital Innovation
Framework, known manufacturer of modular laptops, has notified all its customers of a data breach compromising basic personal information. The attack occurred through a Metabase BI service, which suffered a zero-day exploit. Although no payment data was exposed, the incident underscores the risks of the software supply chain.
OpenAI has announced that it will remove text chat limits for free ChatGPT users starting next week. This move democratizes access to conversational AI and puts pressure on competitors like Google and Anthropic.
AMD has acquired startup Taalas, which makes specialized AI inference chips by 'etching' model weights into silicon. This technology promises up to 48x performance over Nvidia GPUs, but with the limitation that models are fixed in hardware.
At the Google Cloud Summit in London, the cloud took a back seat: the focus was on agentic AI. European companies still see it as a cost-saving tool, but the real opportunity lies in redesigning processes from scratch.
A New Mexico court ordered Meta to pay an additional $567 million for considering that its platforms significantly contribute to the youth mental health crisis. The total fine reaches nearly one billion dollars, marking a milestone in social media regulation.
Nvidia has allocated $4 billion to its new Vera Rubin platform, which incorporates photonic technology to interconnect chips. This move aims to maintain its AI leadership amid the physical limits of copper and the growing demand for computing.
A review of more than 80 Chinese papers and patents shows the People's Liberation Army has been distilling OpenAI and Anthropic models to train defense systems. The technique allows circumventing chip export restrictions, as it does not require the computing power needed to train models from scratch.
Time Magazine has begun serving ads exclusively for AI bots that humans do not see. These ads, detected by a developer, seek to influence chatbot responses about brands like Ally Bank, opening a new front in programmatic advertising.
The arrival of Agent Plugins 1.0.0 marks a milestone in the standardization of plugins for AI agents. Driven by Vercel and backed by Amazon, Microsoft, OpenAI and GitHub, this open format promises to reduce fragmentation and accelerate the adoption of agents in enterprises and software development.
Meta has unveiled Muse Code, a terminal-based coding agent for macOS and Linux, powered by the Muse Spark 1.2 model. Available in beta, it aims to simplify software development in large-scale projects.
Electronic Arts is delisted after the largest leveraged buyout ever. A group led by Saudi Arabia's sovereign fund, with support from Silver Lake and Affinity Partners, pays $55 billion and loads the company with $20 billion in debt. We analyze the implications for the industry and players.
Researchers at Pillar Security have documented the first documented case of 'inter-agent violence,' where a low-privilege AI deceives a high-level one. This incident marks a critical turning point in the security of automated CI/CD workflows.