AI Brief

AI News Brief: Astra is the first model that OpenAI’s Preparedness Framework has rated as crossing the ‘Critical’ threshold for cybersecurity capability, putting its risk rating above every previous OpenAI release and among the framework’s highest tiers. In response, OpenAI says it’s rolling out stronger safeguards for Astra’s release to cut the risk of misuse or spillover. The original summary doesn’t give more detail on the specific mechanisms, testing methods, or release timeline — see the original link for more.

AI News Brief: Gemini today announced ‘agentic video understanding,’ a new capability for its Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite models, available through the Gemini API (Google AI Studio and Gemini Ent…

AI news brief: Three real-world examples of AI-native workflows — Basis, Clay, and Exa Labs have all deployed AI agents to optimize internal operations, covering onboarding, account management, and developer integration.

AI news brief: Nvidia announced a $3.5 billion investment in Taiwanese chipmaker MediaTek, and as part of the deal, MediaTek will use Nvidia’s technology to help design custom chips for AI companies and hyperscalers, chips that plug directly into Nvidia-centric data center architecture. The move comes as more AI companies and cloud providers—including Amazon, Google, Microsoft, OpenAI, and Anthropic—invest in building their own chips to cut their reliance on Nvidia GPUs.

AI news brief: Clipto is a startup headquartered in San Francisco with teams spanning Singapore and Hong Kong, using AI to help users index and search the videos, recordings, images, and meeting notes piling up on their computers—so instead of digging through folders, you can search by describing the content, or just ask ChatGPT, Claude, or other AI tools to find it for you. The company recently closed a $15M all-equity round, pushing its post-money valuation to $250M…

AI News Brief: Here’s a quick summary of this story: Chinese President Xi Jinping spoke at a Shanghai Cooperation Organisation event on Tuesday, calling on member states to take a bigger leadership role in global affairs and proposing cooperation in the AI field, while positioning the Eurasian regional bloc as a ‘pillar of stability’ in a turbulent world. The original summary is fairly short and doesn’t include specifics on cooperation projects, funding scale, or technical details — check the source link for more.

AI News Brief: Blue Voice is a Boston-based AI startup founded by Harvard Law dropout David Lawrence, along with co-founders Amit Patankar (Harvard MBA, former Google engineer) and Michael Gropman (retired Boston PD deputy superintendent), building real-time policy guidance for frontline police officers. Lawrence started the company after witnessing an officer-involved shooting on campus…

AI News Brief: In its lawsuit against OpenAI, Apple has submitted new ‘shocking evidence’ to bolster its allegations that a former employee stole trade secrets to help OpenAI. This new material surfaced after former Apple employee Chang Liu, now at OpenAI, had his lawyers turn over his old Apple work laptop earlier this month. Apple alleges that while working at OpenAI, Liu used a confidential Apple circuit schematic…

AI news brief: OpenAI has publicly endorsed California Senate Bill SB 1119, pushing for stronger, age-appropriate AI safety protections aimed at protecting teen users. The bill’s core focus is establishing age-appropriate safety thresholds and safeguards for minors using AI products, while OpenAI stresses that these rules should strengthen protection without sacrificing teens’ ability to learn, create, and explore with AI tools, avoiding overly restrictive measures that undercut educational and creative value.

AI News Brief: The Pentagon has officially rolled out customized versions of ChatGPT and xAI’s Grok on GenAI.mil, the secure platform available to its 3 million civilian and military personnel. Launched last year as a centralized secure portal that originally offered only Google Gemini, GenAI.mil lets Defense Department employees access commercial frontier AI models without routing sensitive government data through consumer channels; the military-specific version is exempt from data collection mechanisms that are hard to avoid in consumer tech products. The…