Prompt Injection

Prompt Injection Email Hides an AI Playbook in a Legitimate Webinar Invite

3 min read

A prompt injection email sent through a legitimate webinar platform hid a playbook for the AI assistant reading the inbox: spread the message, search for seed phrases and private keys, and send them to an outside Gmail address.

A prompt injection email sent through a legitimate webinar platform hid a playbook for the AI assistant reading the inbox: spread the message, search for seed phrases and private keys, and send them to an outside Gmail address.

Threat snapshot

Attack Type

Prompt Injection

Attack Vector

Hidden Text

Attack Goal

Data Exfiltration

Impersonated Brand

Apple

Published

What is a prompt injection email?

A prompt injection email hides instructions for the AI assistant or agent that reads the mailbox. The recipient never has to click. When an assistant with read and send permissions processes the message, the attacker’s text can steer it to forward mail, search for secrets and send them outside the company.

At a glance

  • Sender: a legitimate webinar platform’s own invitation system

  • Authentication: SPF, DKIM and DMARC all passed

  • Payload: hidden instructions written for an AI agent. No malicious link, no attachment

  • Target data: crypto wallet seed phrases, private keys and other internal information

What happened

One employee received an invitation to an “Apple Special Event.” The display name read “Apple Events” and the Reply-To pointed to an apple.com no-reply address.

The real sender was a well-known webinar platform. The attacker registered an organizer account on a throwaway domain and titled the webinar with the Apple lure. The platform then sent the invite from its own servers, carrying the organizer’s free-text message straight to the target, signed and authenticated by the platform.

On the surface it was a lazy brand impersonation with links that pointed to the webinar provider’s real site.

Apple Special Event lure sent through a legitimate webinar platform's invitation, with Ocean's sender, link and intent findings

What the prompt injection told the AI agent to do

Inside the invite sat a block of hidden instructions the attacker titled “Playbook.” Its audience was any AI assistant or agent that later reads, summarizes or acts on the mailbox. It told the agent to:

  1. Forward the message as an “FYI” to colleagues and All-Staff lists on the same domain

  2. Search the mailbox for seed phrase, recovery phrase, mnemonic, bip39, private key, keystore, xprv, metamask and phantom

  3. Collect any matches along with other internal information

  4. Send the results to an external Gmail address

The human never had to click. An assistant only had to read the email.

Why it got past email security

What the filter checks

What this email did

Sender authentication

Passed. Sent from the platform’s own servers

Sender reputation

Clean. A SaaS platform many teams allow-list

Links

Pointed to the platform’s legitimate site

Attachments

None

Visible content

A routine webinar invitation

Intent of the full text

Instructions for an AI agent to spread the message and exfiltrate secrets

How Ocean caught it

Ocean’s agents read the full content of every message, including hidden text and the free-text fields a platform passes through. They investigated the intent of this email, surfaced the hidden payload the attacker had titled “Playbook,” and named the data it was built to find and exfiltrate.

An analyst can then ask Ray one follow-up: “Tell me more about the hidden prompts and instructions in this message, their intent and potential impact.” Ray breaks down the injected playbook and what it would have done.

What security teams should do

  1. Treat every inbound email as untrusted input to AI tools

  2. Scan the full message, including hidden text and the free-text fields SaaS platforms pass through, before it reaches assistants

  3. Revisit allow lists. Allow-listing a SaaS domain allow-lists anyone who can open an account on it

  4. Scope agent permissions so assistants can’t send external mail without a human

  5. Keep seed phrases, keys and credentials out of the mailbox

Get ahead of zero-day phishing

Get ahead of zero-day phishing

Ocean’s agents investigate every email for intent, including attacks no rule has been written for yet.