Google Home Cameras Start Using Clothing and Sounds to Explain What Happened
Google's June 23 Google Home update lets Gemini camera features use additional cues such as clothing when a face is not visible and add sounds like alarms, footsteps, glass breaking, and barking to event descriptions. The change makes Nest camera alerts more useful, but it also makes privacy and subscription settings more important.
Google’s AI Talent Losses Put Coding and Science Roadmaps Under Pressure
Google DeepMind lost Noam Shazeer to OpenAI and John Jumper to Anthropic in the same week. The moves matter because AI labs are competing for researchers who can steer coding agents, scientific AI, and frontier model strategy, not just write papers.
ByteDance Seedance 2.5 Pushes AI Video Toward Production Workflows
ByteDance’s Seedance 2.5 is expected in early July with 30-second native 4K video, up to 50 multimodal reference inputs, and tighter editing controls. The launch matters less as a demo-war milestone than as a sign that AI video tools are being built for repeatable production workflows.
Claude Tag Turns Slack Channels Into Shared AI Workspaces
Anthropic’s Claude Tag puts a shared, permission-scoped Claude inside Slack channels for Claude Enterprise and Team customers. The launch moves workplace AI from private chatbot sessions toward visible, persistent agents that can remember channel context, use approved tools, and work asynchronously.
Superhuman’s GPTZero Deal Brings AI Detection Into Daily Writing Tools
Superhuman is buying GPTZero, putting AI detection, authorship checks, hallucination detection, plagiarism review, and AI image detection closer to everyday writing workflows. The deal shows that content authenticity is becoming part of productivity software, not just a classroom policing tool.
Google DeepMind’s A24 Deal Puts AI Inside the Filmmaking Workflow
Google DeepMind and A24 are launching a multi-project AI research partnership, with Google reportedly investing about $75 million in the studio. The deal is less about one AI movie tool than a test of whether creative AI can be shaped inside real film workflows without handing a tech company the studio library.
LastPass Says Klue Breach Exposed Support Case Data, Not Password Vaults
LastPass says attackers used Klue-held OAuth tokens to access customer CRM and support case data in Salesforce, while its password vaults and core infrastructure were not affected. The practical risk is targeted phishing and social engineering built from real support histories.
CISA’s June 23 Deadline Puts Cisco SD-WAN, Chrome, and Arista EOS on the Triage List
CISA’s June 23 remediation deadline covers three actively exploited flaws across Cisco Catalyst SD-WAN Manager, Google Chrome’s V8 engine, and Arista EOS. The useful move for security teams is not treating them as one patch chore, but triaging each layer: network control plane, browsers, and tunnel decapsulation paths.
OpenAI Codex Record & Replay Turns Workflow Demos Into Reusable Skills
OpenAI’s new Codex Record & Replay feature lets eligible macOS users demonstrate a repeatable workflow once and turn it into a reusable skill. It could make desktop and browser automation easier to capture, but teams need to treat recordings, permissions, and sensitive data carefully.
NVIDIA Rubin Pushes AI Data Centers Toward Hotter, Drier Cooling
NVIDIA says its Rubin-generation AI infrastructure can run fully liquid-cooled servers with 45°C coolant, cutting facility cooling water use from conventional tower-based levels to near zero in favorable climates. The design is a real shift for AI factories, but it does not erase the water tied to power generation, chip manufacturing, or local data center siting fights.