According to OpenAI on X, Codex Automations can now run in the same thread, preserving full conversational context to resume work without rehydration. As reported by OpenAI, the system can schedule future tasks and automatically wake to continue long-running jobs, enabling workflows like closing open PRs, task follow-ups, and monitoring fast-moving conversations. According to OpenAI, this reduces context-loss friction in agentic workflows, improving reliability for software delivery pipelines, customer support escalations, and sales engagement cadences. As reported by OpenAI, the persistent thread model also supports timed triggers and stateful execution, creating opportunities for AI ops, DevEx automation, and hands-off backlog grooming.

According to @godofprompt citing @claudeai on X, Anthropic’s Claude Cowork now supports scheduled tasks that automatically run recurring workflows such as morning briefs, weekly spreadsheet updates, and Friday team presentations. As reported by the @claudeai post, the feature enables time-based automation so teams can operationalize AI agents for routine knowledge summaries, data refreshes, and content preparation without manual prompts. According to the X video by @claudeai, this expands Claude’s role from on-demand assistant to a lightweight automation layer, creating opportunities to standardize reporting cadences, reduce coordination overhead, and integrate with calendars and data pipelines. For enterprises, the business impact—per the @claudeai announcement—is higher utilization of AI in back-office operations, measurable time savings on recurring tasks, and more consistent outputs for stakeholders.