Curated AI news from the last 24 hours — models, agents, enterprise, and developer tools.
Anthropic reportedly told investors its third-quarter revenue reached about $16.8 billion, up roughly 45% from the prior quarter. People familiar with the matter say it also reached adjusted profitability for the first time. The company is preparing for a potential IPO, with reports pointing to a November listing.
cnbc.com ↗SiliconANGLE reports Anthropic disclosed 45% sequential revenue growth to $16.8 billion in Q3 as it prepares to go public. The figures follow a leaked confidential filing in September that showed rapid growth alongside a large 2025 operating loss.
siliconangle.com ↗Security startup Xbow reportedly used OpenAI's GPT-5.6 Sol to chain three Chrome zero-days into a sandbox-escaping exploit on Windows. Four researchers spent six days and about $40,000 in tokens; Google patched the bugs and paid a $250,000 bounty. The account rests on Xbow's own description.
theregister.com ↗Shares of Oracle and several AI cloud providers fell on concerns about CoreWeave's debt. The drop highlights investor sensitivity to the financing behind AI data center buildouts.
siliconangle.com ↗Google has rolled out Gemini 4 Argon to Gemini app Pro and Ultra subscribers, after an earlier release to vetted cybersecurity defenders. The rollout adds a new tasks mode. Google's chart from the model's unveiling showed it ahead of Anthropic's Claude Opus 5.5 on DeepSWE v1.1.
siliconangle.com ↗Microsoft plans to run most of its AI inference workloads on its own Maia accelerators, starting with the third-generation chip. The move is intended to reduce reliance on Nvidia and lower costs.
cnbc.com ↗Grok Voice 3.0 reportedly outscores OpenAI and Google models on a speech-to-speech benchmark index. The release is positioned as a step forward for xAI's voice capabilities.
siliconangle.com ↗