Cursor upgraded Cloud Agents and its agent harness with event-driven subscriptions to PRs and Slack threads, a /goal command for long-lived objectives, subagents that run on their own isolated virtual machines, and smoother mid-task steering.
Kuaishou's Q2 2026 results, released August 19, show Kling AI revenue above RMB 850 million with more than 200% year-over-year growth, alongside the launch of Kling MCP and Kling CLI for agent-driven batch video creation and native 4K output in the Kling 3.0 series.
Perplexity published a technical deep-dive on Brain, the memory system for its Computer agent, detailing its Markdown-filesystem architecture, offline 'Dream' agents that refine memory overnight, and 30-day production results showing a 9.3-point correctness gain.
OpenAI is reaffirming Zero Data Retention for eligible API customers and previewing Private Safety Processing, a new system that detects misuse patterns across related interactions without exposing content to OpenAI staff, in testing now with a September rollout planned.
Replit is powering its Free Mode with GPT-5.6 Luna, giving free users fast, no-usage-cost AI coding assistance that escalates to GPT-5.6 Sol for harder reasoning tasks while preserving full project context.
Waymo has integrated Google's Gemini assistant into its custom-built Ojai autonomous vehicles, letting riders control cabin features, search nearby businesses, and ask general questions by voice, independent of the self-driving system.
Perplexity's Computer agent platform can now be triggered by sending, forwarding, or cc'ing computer@perplexity.com, letting it run full tasks like building financial models or summarizing legal redlines without leaving your inbox.
OpenAI is expanding ChatGPT Ads to 31 European countries including Germany, France, Spain, Italy, and the Nordics, its largest expansion since the February U.S. pilot, with ads shown only to Free and Go tier users.
OpenAI is committing $5 million and a year of work to help government oversight bodies keep pace with AI-assisted national security decisions, built around three principles: human judgment primacy, traceability, and empowering oversight institutions.