An invitation email sent from xAI’s own address, noreply@x.ai, carried a hidden phishing link, according to Wes Morrill, who posted it on X on 14 October 2026. He wrote that it passed SPF, DKIM and DMARC and had the blue check, and later that “it really was sent by xAI.”
Elon Musk quoted the post: “Grok @Bot can detect scam emails.” Five days earlier, on 9 October, an X product engineer had replied to another user who posted the same invite: “This was a system exploit. It’s fixed. Don’t click the link.”
Business Pill · GUARDRAILS
A one-minute explainer of guardrails: the rules enforced around an AI model. It teaches the general idea only and says nothing about any company in this story.
The key insight: As we read it, the email passed every check because, on Morrill’s account, it was genuine: the sender was xAI, and the phishing link came in through a team name that xAI’s own template did not escape.
The screenshot in Morrill’s post shows an email from “SpaceXAI” at noreply@x.ai with a blue verified check. It reads “You have been invited to join the X Founders Private Program,” followed by an “Accept Invitation” link and the sign-off “- xAI Team.”
Morrill wrote that “something felt off,” so he asked Grok Bot, xAI’s agent product, to do a security review. He wrote that it found a hidden phishing link, explained why every safety check passed and wrote a report to xAI, including a proposed fix.
Cybernews, which covered the post on 15 October, describes Morrill as the lead Cybertruck engineer at Tesla. His post had 672,577 views when we read it through the fxtwitter API at 14:54 UTC on 15 October.
Checks the SpaceXAI invite email passed, per Wes Morrill’s 14 October 2026 post (SPF, DKIM, DMARC, blue check), and the step he says failed: the attacker’s team name went into xAI’s invite template as raw HTML without escaping.
What Was Posted on X
An email from noreply@x.ai landed in my inbox, passed SPF, DKIM, DMARC, and had the blue check
Something felt off, so I asked @bot to do a security review. It found a hidden phishing link, explained why every safety check passed, wrote a report to xAI – including proposed fix!
Wes Morrill posted the invite email on 14 October 2026, writing that it came from noreply@x.ai, passed SPF, DKIM and DMARC and had the blue check, and that Grok Bot found a hidden phishing link. His replies in the thread, quoted above, explain the team-name mechanism and the fix.
Elon Musk quoted Morrill’s post 26 minutes later with one line: “Grok @Bot can detect scam emails.” The piece reports it as Musk’s framing of the episode.
@sonyasupposedly @spacexai This was a system exploit. Itโs fixed. Donโt click the link. Thanks.
Mridul Singhai, whose X profile reads “Product Engineering @X”, replied on 9 October 2026 to an earlier post of the same invite: “This was a system exploit. It’s fixed. Don’t click the link. Thanks.” He posted it from his own account; the piece does not treat it as a company statement.
How It Passed the Checks
Morrill explained the mechanism in a reply in the same thread, describing what Grok Bot found: “the attacker named their xAI team with raw HTML which xAI’s invite template dropped into the email without escaping it.”
That, he wrote, is how a fake “Accept invitation” button appeared in a genuine email: “That’s why every authentication check passed – it really was sent by xAI.”
As we read it, that is the point of the story: on Morrill’s account, the checks confirmed who sent the email, and the sender was real. The problem sat in what xAI’s own template let a team name put inside it.
The Fix and the Earlier Report
In a second reply, Morrill described the fix as Grok Bot proposed it: “HTML-encode team names when they go into emails, and restrict what characters a team name can contain.” He added: “Looks like that feedback is now implemented.”
The same invite had been posted on X before. On 9 October 2026 at 18:37 UTC, the account @sonyasupposedly shared a screenshot of an email with the subject “You’ve been invited to a team on SpaceXAI” and the same X Founders Private Program text, asking: “is this actually from @spacexai because it feels like phishing.”
Mridul Singhai, whose X profile reads “Product Engineering @X,” replied to that post and to @spacexai at 21:13 UTC the same day: “This was a system exploit. It’s fixed. Don’t click the link. Thanks.” He linked that reply under Morrill’s post on 14 October.
Morrill’s posts do not say when he received his email, so they do not show whether it arrived before or after Singhai’s 9 October reply.
Where Grok Bot Fits
Musk’s quote post presented the episode as a Grok capability. Grok Bot is the agent product of SpaceXAI, and on 9 October 2026 the official @bot account announced that Grok Bot had its own email address, as we reported that day.
Morrill wrote that he also asked his bot to add a security routine that checks his emails for anything phishy, and that it has since flagged a few more as “Likely Phishing,” which he called “more obvious.”
Cybernews raised a caveat about the same agent. It quotes SpaceXAI as stating: “The computer is assigned to your user account, not an individual Bot. Do not place a credential or file on it if another Bot on your account should not be able to use it.”
Cybernews information security researcher Rasa Jurgutytฤ told the outlet that Grok Bot could become a “target for prompt injection and other attacks, since there’s a possibility that it could handle financial transactions that could be manipulated.”
The Structural Read
As we read it, SPF, DKIM and DMARC did what they are built for here: Morrill says the email really was sent by xAI, and the problem sat inside the message, in what the invite template accepted.
As we read it, the episode puts an agent on both sides of the inbox: Grok Bot, which on 9 October got its own email address, is the agent Morrill says caught an email sent by its own maker.
As we read it, the fix Morrill describes is a template change, not an email-security one: HTML-encode team names and restrict the characters they can contain.
Wes Morrill, on X, 14 October 2026
“To users: Reminder not to click links on unexpected emails, even if the sender checks out”
Three Implications
USERS Morrill’s own advice in the thread: do not click links in unexpected emails, even when the sender checks out.
XAI Singhai wrote on 9 October that the exploit was fixed; Morrill wrote on 14 October that the proposed fix looked implemented.
AGENTS IN THE INBOX Cybernews quotes SpaceXAI as saying a Grok Bot computer is assigned to the user account, not to an individual bot.
The framework argues that “the limiting factors are organizational readiness, governance, workflow redesign, training, and trust.”
As we read it, this episode tests the last of those: an email can pass every authentication check and still carry a phishing link, and Musk’s post presents an agent as the check; the framework does not discuss xAI’s invite emails.
The two changes as Wes Morrill described them on X on 14 October 2026; he wrote that the feedback looked implemented, and X engineer Mridul Singhai wrote on 9 October that the exploit was fixed.
What Is Not Established
The posts and Cybernews’ article do not include a statement from xAI or SpaceXAI beyond Singhai’s reply, which he posted from his own account.
They do not say how many people received the invite, what the link led to, whether anyone entered details, or when the fix Morrill describes went live.
Business Engineer Framework
The Agentic Harness War
A Business Engineer framework on how agents move from tools people consult to systems people deploy.
Two X users posted the same SpaceXAI invite from noreply@x.ai, on 9 and 14 October. Morrill wrote that it passed SPF, DKIM and DMARC because an attacker’s team name, written in raw HTML, went into xAI’s own invite template without escaping.
An X product engineer wrote on 9 October that the exploit was fixed, and Morrill wrote on 14 October that the proposed fix looked implemented. Musk used the post to say Grok can detect scam emails.
A note on sourcing. We read Wes Morrill’s 14 October 2026 post on X and his three replies in that thread, Elon Musk’s quote post, the 9 October post by @sonyasupposedly and Mridul Singhai’s reply in full, and looked at both email screenshots; we do not reproduce the recipients’ addresses. We read Cybernews’ 15 October article in full. We did not test the link or contact xAI. The Grok Bot email-address context comes from our own earlier reporting.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur, HackerNoon and Search Engine People. Find him on LinkedIn and Substack. See how we source.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
