AI DAILY
Artificial Intelligence Daily Digest
AI Helps Find Major Security Bug in Java, a Platform Vital to the Internet
AI helped uncover a major security vulnerability in Java, a software platform critical to how the internet runs, according to The New York Times.
Microsoft unveils six specialized AI models at Ignite, MAI Code 2 at about 5% cost
At Ignite, Microsoft introduced six specialized AI models, saying its in-house MAI Code 2 approaches OpenAI and Anthropic models in coding at about 5% of the cost, plus models for cybersecurity and AI agent judgment tasks.
OpenAI launches GPT-6.2 Astra and Sol with lower API prices and better coding
OpenAI released GPT-6.2 Astra and Sol, citing better professional outputs, coding, computer use and efficiency at lower cost, with some benchmark gains and lower API prices.
US Introduces New Restrictions on AI Chip Exports
According to Germany's WirtschaftsWoche, the US has introduced new restrictions on AI chip exports, tightening export controls on the overseas supply of AI chips.
Zhipu MaaS platform tops 10 million developers, GLM open models over 60% of tokens
Zhipu's MaaS platform has surpassed 10 million developers, with its open-source GLM models accounting for more than 60% of token calls.
Anthropic releases Claude Opus 5.5, citing 40% lower cost than Opus 5
Anthropic released Claude Opus 5.5, saying it leads in agentic coding and knowledge work and costs 40% less to run than Opus 5 on typical workloads.