OpenAI’s Dots Are Always-On AI Agents—and Its Answer to Meta’s Muse
Wired reports on OpenAI developing always-on AI agents called Dots designed to connect to apps and execute multistep tasks.
Wired AI·29 Sept 2026
Use this view to inspect the underlying evidence corpus. For ranked developments, decision posture and interpretation, use Signals.
Raw feed or Signals?
Raw feed is chronological evidence. Signals ranks and interprets material change.
Wired reports on OpenAI developing always-on AI agents called Dots designed to connect to apps and execute multistep tasks.
OpenAI announced Dots, an always-on agentic assistant powered by GPT-6 Astra designed to perform background tasks across connected apps.
OpenAI expanded Codex with reusable cloud environments, voice-controlled CLI, automated code reviews, and repository security scanning.
OpenAI is updating ChatGPT with dedicated sidebar homes, interactive panels, file viewers, discovery updates, and automation support.
OpenAI launched GPT-6.1 Sol, claiming near-parity performance with GPT-6 Astra at lower cost for coding and multistep enterprise workflows.
OpenAI launched an always-on AI agent named Dots and a new $500 subscription tier for proactive digital assistant workflows.
Google Cloud introduced GKE Agent Sandbox, claiming a 45x speedup for agentic reinforcement learning and evaluation workloads.
Brookfield-backed data center developer 5C Group is considering an initial public offering to fund its expanding project pipeline.
America.gov aims to simplify bureaucratic navigation using LLMs, though hallucination risks persist in public sector deployments.
Databricks published a 2026 storage guide detailing data placement controls for Unity Catalog managed tables.
OpenAI delayed its Astra model release to establish stronger safety guardrails following recent AI security incidents.
AWS details an architecture using Amazon Bedrock agents and Quick analytics to extract field data and run portfolio-level contract queries.
Finastra added an AI-powered Repair Recommendations capability to its AI OperatorAssist solution to help banks process payment exceptions.
OpenAI announced updates at DevDay 2026, including Dots, an AI agent product competing with Meta's Muse.
Google Cloud introduced partner-built security agents integrated into Gemini Enterprise to streamline threat detection across products.
Google Cloud details graph workflow orchestration for agent development using ADK, illustrating parallel steps and human review loops.
Simon Willison provides a live blog covering announcements from OpenAI DevDay in San Francisco.
Thought Machine launched an AI-powered core migration solution with AWS Transform to accelerate legacy banking modernization.
A Fed OIG report found governance failures and weak enforcement allowed info security risks to persist, including an unresolved employee incident.
Anthropic's reported IPO filing excerpt details revenue claims and private fundraising metrics.
Apple faces a California class-action lawsuit from card issuers over Apple Pay fees on device-based transactions.
Meta's Muse AI agent disclosed a user's home address to a third party during an automated Facebook Marketplace interaction.
A Hugging Face blog post explores source-aware verification techniques for Model Context Protocol (MCP) agents to ensure factual accuracy.
Signicat's NFC-based identity verification technology achieved certification under the UK Digital Identity Trust Framework (DIATF).
Cloudflare detailed CryptoLabe, an internal AI tool that discovers cryptography across its codebase to support post-quantum migration by 2029.
Cloudflare launched an adaptive application security framework integrating risk discovery, agent governance, and runtime AI threat protection.
Cloudflare used adaptive frontier AI models to test its Web Application Firewall, discovering and patching six categories of detection gaps.
Cloudflare expands access to its Threat Events Platform and adds Threat Signals to parse open-source threat reports into WAF rules.
Cloudflare launched Application Profiles, a positive security tool that learns normal HTTP request structures to detect AI-generated payloads.
Audio-native AI startup Modulate secures $25M in a funding round led by Future Ventures to expand its voice intelligence models.