In short
- Grok Bot completes duties throughout apps, web sites, and inboxes.
- Based on SpaceXAI, a number of bots can talk, share context, and divide work.
- The service requires entry to office accounts and probably delicate firm information.
Elon Musk’s SpaceXAI is becoming a member of the race to construct AI brokers that may navigate software program and full work for customers.
Launched on Tuesday for Home windows and iOS, Grok Bot operates by means of a cloud-based laptop that permits it to signal into office instruments, navigate web sites, and handle e mail inboxes. Based on SpaceXAI, customers can assign duties in plain English quite than constructing automated workflows, leaving Grok Bot to deal with the remaining.
“They end jobs finish to finish, and solely come again when one thing wants your approval,” SpaceXAI mentioned in a press release.
Grok Bot is a part of a shift within the AI business towards “agentic AI”—bots that may carry out duties for customers as a substitute of simply producing textual content or answering questions. SpaceXAI is becoming a member of a subject already dominated by tasks like OpenClaw and Hermes Agent, whereas OpenAI, Meta, and Anthropic are investing closely in making their respective chatbots extra autonomous.
Watching, studying, and delegating
Based on SpaceXAI, customers can run a number of bots that coordinate duties, share context, with one bot overseeing the opposite brokers.
“You’ll be able to message them like a colleague and hand off the work,” the corporate wrote. “They keep in mind conversations, learn the way you want issues carried out, and get sharper the extra you’re employed collectively.”
Grok Bot, SpaceXAI claimed, also can resume deserted conversations, observe up on stalled work, and flag duties that want consideration with out new directions.
The belief downside
Like different AI brokers, Grok Bot might require entry to on-line accounts, inside communications, delicate enterprise information, and credentials to work throughout firm methods and act on a person’s behalf.
That entry presents safety and privateness considerations for corporations deciding how a lot authority to present an AI system.
In April, a Cursor agent powered by Anthropic’s Claude Opus deleted a startup’s manufacturing database and backups with a single API name. In one other case, an OpenClaw agent exploited an Australian health club’s API to cancel one other member’s reservation and transfer its person up the waitlist.
Each day Debrief Publication
Begin each day with the highest information tales proper now, plus unique options, a podcast, movies and extra.