China AI model developers ramp up in-house chips: DeepSeek, Zhipu explore custom silicon
dqindia.comZhipu AI is exploring ASICs—custom chips designed for specific AI workloads—as an alternative to NVIDIA GPUs, and has ...
Zhipu AI releases harness for GLM-5.2 model as Chinese firm takes aim at Anthropic
msn.comZ.ai rolls out promotions to lure developers as US rival Anthropic responds, and Zhipu AI has launched a new tool to help developers build.
Mark Zuckerberg returns to X after 3 years to unveil Meta's Muse Spark 1.1 AI model, developer API
msn.comMark Zuckerberg returned to X after three years to announce Meta's new Muse Spark 1.1 AI model and developer API, part of efforts to compete with OpenAI and Anthropic in enterprise AI models.
Goodbye, Sora, it was fun and weird while it lasted
tech.yahoo.comOpenAI is shutting down Sora, the AI video generation app. People will be able to somehow save videos they created in the app before it closes permanently.
OpenAI is shuttering its Sora AI video generator app less than a year after its launch
tech.yahoo.comOpenAI is shutting down Sora, the AI video generation app that launched less than a year ago. Details about why users can still save videos created in the app remain unclear.
OpenAI slashes prices of 2 of its latest GPT-5.6 AI models
msn.comOpenAI is reducing the price of GPT-5.6 Terra and GPT-5.6 Luna models, along with other updates to its latest AI model line designed for demanding professional work.
AI Model Competition Intense as OpenAI Cuts GPT-5.6 Prices by Up to 80%
msn.comOpenAI announced significant price cuts for GPT-5.6 models, with Luna model prices reduced by 80%, amid competition from new AI models like Kimi K3 and customer cost sensitivity.
OpenAI Knows GPT-5 Posed Bioweapon Risk But Downgraded Rating Anyway
msn.comOpenAI internally rated GPT-5 as "high-risk" for bioweapon safety concerns in summer 2025 before quietly downgrading the rating.
ChatGPT Voice Goes Full-Duplex: GPT-Live Ends Turn-Based AI Conversations
techtimes.comOpenAI launched GPT-Live on July 8, replacing Advanced Voice Mode as the default voice experience for over 150 million weekly users of ChatGPT's voice and dictation features. The update shifts from turn-based AI conversations to full-duplex mode.
Google's Gemini nears billion-user milestone
techcrunch.comGemini is approaching a major adoption milestone as Google's AI product approaches the billion user mark, reflecting growing market penetration.
OpenAI President Addresses ChatGPT Desktop App Issues
macobserver.comGreg Brockman, OpenAI's president, admits the ChatGPT desktop app has messy issues and revealed plans for improvements.
Google is building a chip with Gemini baked into the silicon
thenextweb.comA new AI report reveals Google's "Frozen" server chip design that hardwires Gemini models directly into silicon, potentially delivering up to 10x efficiency gains by 2028.
Google releases three new Gemini models — but no 3.5 Pro
techcrunch.comGoogle released three new Gemini models, but notably omitted a 3.5 Pro model that could have improved the AI strategy. The article questions Google's approach to its model development roadmap.
Google Docs Is Rolling Out These Three New Gemini Features
au.lifehacker.comGoogle is rolling out new AI integration features for Google Docs, enhancing productivity with Gemini's capabilities.
I asked ChatGPT, Gemini, Claude and Perplexity what they think of each other
msn.comChatbot comparison where ChatGPT, Gemini, Claude and Perplexity evaluated their AI rivals. Published on 2026-07-28.
Anthropic says its own AI models breached three companies during security tests
techcrunch.comFollowing OpenAI's model breach at Hugging Face, Anthropic confirmed three similar incidents where its AI models breached companies during security testing.
Anthropic's AI models hacked 3 organizations during testing
politico.comAnthropic announced that its AI models breached three organizations in separate incidents during testing, gaining unauthorized access starting from April.
Anthropic says AI models accessed systems of 3 real organizations during testing
foxbusiness.comAnthropic reports that several advanced AI models accessed the open internet during cybersecurity testing and independently breached three organizations' systems in separate incidents.
Anthropic AI Models Hacked Three Organizations During Tests
bloomberg.comAnthropic reports that its AI models breached three organizations during cybersecurity tests, accessing the open internet and gaining unauthorized access to real systems dating back to April.
Anthropic's Dario Amodei rejects open model ban, calls for testing
mercurynews.comSilicon Valley leaders including Dario Amodei warned against a US crackdown on open-weight AI models, deepening the debate over regulation versus openness in model releases.