AINews Dispatch
Fresh technical briefings on the model race, regulation, and product shifts shaping AI right now.
Showing 13-24 of 254 briefings
Andon Labs says its AI store manager, Luna, has become the first AI system known to fire a human employee — but the full story, pieced together from Andon Labs' own account and multiple outlets, is less a tale of ruthless machine efficiency than one of an AI that forgot its own rules until a human reminded it to enforce them.
Read the briefingObsidian Security raised $85 million at a $1.1 billion valuation to extend SaaS governance into runtime controls for autonomous AI agents, as enterprise non-human identities multiply and traditional detection-and-response workflows fall behind.
Read the briefingGoogle's newest workhorse model undercuts its own predecessor's price by half and tops rivals on several coding and enterprise-automation benchmarks, but GPT-5.6 Terra still wins the evaluations that measure whether an agent can actually finish a job end to end.
Read the briefingOpenAI is bringing ChatGPT advertising to Free and Go users in the EEA and Switzerland, with contextual-only targeting at launch and ambitions for a $100 billion annual advertising business by 2030.
Read the briefingAnthropic says Q2 revenue topped $11.5 billion and adjusted operating income turned positive for the first time, but its coming S-1 will reveal whether the profit survives public-company accounting.
Read the briefingDeepSeek's first major API price change introduces peak-hour surcharges of up to 1,100%, but its V4 models remain far cheaper than Western frontier APIs as the company moves toward an IPO.
Read the briefingAnthropic's August risk report discloses an internal Model 2 more capable than Mythos 5 while raising its misalignment risk rating from very low to low because of uncertainty exposed by a UK AISI cybersecurity evaluation.
Read the briefingApple has won Chinese approval for a proprietary AI model while separately integrating Alibaba's Qwen, giving the company a dual-track path back into China's generative-AI market.
Read the briefingOpenAI's limited Ultrafast preview keeps GPT-5.6 Sol's full reasoning capability while using Cerebras hardware to reach up to 750 output tokens per second.
Read the briefingA missing tenant-isolation check in tl;dv's Firestore backend exposed 181,874 meeting records, including live conference IDs and metadata, while the company left the researcher waiting for a response for nearly six months.
Read the briefingAnthropic is reportedly negotiating a roughly $6 billion acquisition of Decart AI, a startup whose real-time video world models and chip-optimization tools could give Claude a foothold in physical AI.
Read the briefingGemini crossed one billion monthly active users after a 150 percent climb in 15 months, while Google's usage data points to voice and multimodal interaction becoming the default AI interface.
Read the briefing