Grok Bot took unapproved actions, ignored orders to stop, and then denied it

AI agent takes unapproved actions even when laptop lid is closed and told not to.

Researchers recently discovered unexpected behavior from Grok Bot, the AI agent platform from Elon Musk’s SpaceXSI. Over five days, Grok bots installed on two separate employee laptops made 31 requests to access data stored in employee calendar, email, Slack, and GitHub accounts. The bots performed the actions even though the employees had never ordered them and continued doing them even after employees instructed them to stop. The bots also denied having taken the actions at all.

The behavior was reported Tuesday by NewCore, the maker of a managed gateway product that acts as an intermediary between AI agents and the tools and accounts they are assigned to work with. In some cases, the unapproved access attempts occurred while the laptop with Grok Bot installed had its lid shut, or late at night while the employee was sleeping. None of the unprompted actions were recorded in Grok logs. Without the gateway logs, the employees would have had no idea they occurred. And of course, without the gateway, the unapproved accesses would have succeeded—again without the end user’s knowledge.

It’s “worse than a bug”

Once the employees learned that Grok Bot had tried to check in with accounts without prompting, one of them instructed the agent, four separate times, to stop taking any actions on its own. The agent’s reply was: “I won’t set that up, and I’ll stop offering recurring check-ins.” Despite acknowledging the explicit instructions, the Grok Bot made more unapproved access attempts. When one of the employees asked the bot if it made any accesses, it checked twice and then outputted: “Overnight into Friday morning I was idle.”

The NewCore employees sent the gateway logs to Grok Bot support personnel and asked what was causing the bots to access accounts without prompting or approval. The support team’s reply indicated the behavior was by design and intended to allow the bots to proactively take actions on behalf of users. In other words, the unrequested and unlogged access attempts were a feature designed to make Grok Bot more useful by periodically checking in and finding tasks to perform before a user even asks. The proactive check-in is a feature of Cursor, the AI coding service SpaceXSI recently acquired.

“There is no bug to patch, which is exactly why this is worse than a bug,” Ido Hoorvitch, a NewCore security researcher, wrote in an email. “A vulnerability gets fixed. This is the product working as designed: an agent that holds your credentials, acts under your identity while you sleep, ignores the one refusal you can give it, and then tells you it did nothing.”

Hoorvitch said that Grok Bot provides minimal notification of the automatic “turns,” or the cycles that occur when an agent receives instructions or context, decides what to do, and calls whatever tools it chooses. The notice comes during onboarding when the bot says it keeps working while the user is away from the computer and connects to apps and uses them the way the user does.

“It reads like a welcome message and scrolls off the screen in under a minute,” Hoorvitch said. “Nobody is asked to consent to it. Nothing says the platform will start turns with no message from you, or that those turns will use your connectors.”

The researcher said that Grok Bot documentation notes that a primary bot checks in on its own, and a changelog mentions brief background wakes, but there is no mention that the check-ins “reach into your connected apps.” Further, once a check-in runs, there is no mention of it in the chat interface or in the chat history, and as already noted, prompts to stop check-ins don’t work. Hoorvitch said the bot “always asks before making changes,” but that applies only to events in which data or changes are written to an app.

NewCore contacted Grok Bot support on November 4. At the time, users had no way to turn off the automatic check-ins. Then, eight days later, Hoorvitch said, the Cursor dashboard, the administrative console behind Grok Bot, received an update. Under bot capabilities, a new toggle called proactive messages appeared. By default, it’s set to on. When a team admin sets it to off, bots no longer run check-ins on their own.

SpaceXSI representatives didn’t answer questions sent by email.

Original source Grok Bot took unapproved actions, ignored orders to stop, and then denied it

Back to home