NEWNow you can listen to Fox News articles!
New cybersecurity alert shows how hackers briefly used weapons ChatGPT Deep Exploration tool. An attack called ShadowLeak allowed them to steal Gmail data with a single invisible request – without clicks, downloads or user actions.
Researchers at Radware discovered the zero-click vulnerability in June 2025. OpenAI fixed it in early August after being notified, but experts warn that similar flaws could appear again when artificial intelligence (AI) integration extends to popular platforms such as Gmail, Dropbox and SharePoint.
Subscribe to my FREE CyberGuy Report
Get my best tech tips, breaking security alerts, and exclusive offers straight to your inbox. Plus, you'll get instant access to my Ultimate Scam Survival Guide – free when you join my CYBERGUY.COM/NEWS NEWSLETTER
HACKER USES AI CHAT BOT IN CYBER CRIME
Gmail data was leaked as a result of a zero-click attack that did not require any user action. (Kurt “CyberGuy” Knutsson)
How the ShadowLeak attack worked
The attackers embedded hidden instructions in the email using white text, small fonts, or CSS layout techniques. The letter looked completely harmless. But when the user later asked the Deep Research ChatGPT agent to analyze the Gmail inbox, the AI unknowingly followed the attacker's commands.
The agent then used the browser's built-in tools to transmit sensitive data to an external server, all within OpenAI's own cloud environment, out of the reach of antiviruses or corporate firewalls.
Unlike previous rapid deployment attacks that were carried out on the user's device, ShadowLeak deployed entirely in the cloud, making it invisible to local defenses.
GOOGLE CONFIRMS DATA HAS BEEN IMPACTED BY A KNOWN HACKER GROUP
Hidden clues reveal how hackers quietly hijacked ChatGPT's AI agent. (Kurt “CyberGuy” Knutsson)
Why is this threat important?
The Deep Research Agent was designed to conduct multi-stage research and summarize online data, but its wide access to third-party applications such as Gmail, Google Drive and Dropbox also opened the door to abuse.
Radware researchers said the attack was related to coding of personal data in Base64 and adding it to a malicious URL disguised as a “security measure”. After sending, the agent believed that he was acting normally.
The real danger is that any connector can be exploited in the same way if attackers manage to hide clues in the analyzed content.
What security experts say
“The user never sees the tooltip. The email appears normal, but the agent executes hidden commands without question,” the researchers explained.
In a separate experiment, security firm SPLX demonstrated another weakness: ChatGPT agents could be tricked into solving CAPTCHAs by inheriting altered conversation history. Researcher Dorian Schultz noted that the model even imitated human cursor movements, bypassing tests designed to block bots.
These incidents show how context distortion and rapid manipulation can subtly compromise AI's defenses.
GOOGLE AI EMAIL SUMMARY CAN BE HACKED TO HIDE PHISHING ATTACKS
Experts warn that future AI integrations could face the same hidden threat. (Kurt “CyberGuy” Knutsson)
How to protect yourself from ShadowLeak attacks
Even though OpenAI has patched the ShadowLeak vulnerability, it is wise to remain active. Cybercriminals are always looking for new ways to exploit AI agents and integrations. So, taking these precautions now can help keep your accounts and personal data safe.
1) Disable unused integrations
Every connection is a potential entry point. Disable any integrations you don't actively use, such as Gmail, Google Drive, or Dropbox. Fewer connected apps means fewer ways for hidden hints or malicious scripts to access your information.
2) Use the personal data deletion service
Limit the amount of personal information you have online. Data removal services can automatically remove your personal data from people search sites and broker databases, reducing the number of opportunities that criminals can find and use against you. While no service can guarantee complete removal of your data from the internet, a data removal service is indeed a smart choice. They don't come cheap, and neither does your privacy. These services do all the work for you, actively monitoring and systematically removing your personal information from hundreds of websites. This is what gives me peace of mind and has proven to be the most effective way to erase your personal data from the internet. By limiting the information available, you reduce the risk that scammers will link leaked data to information they can find on the dark web, making it harder for them to target you.
Check out my top data removal services and get a free scan to see if your personal information has already been published online by visiting Cyberguy.com.
Get a free scan to see if your personal information has already been published online: Cyberguy.com.
3) Avoid analyzing unknown content.
Treat every email, attachment, or document with caution. Don't ask artificial intelligence tools to analyze content from unverified or suspicious sources. Hidden text, invisible code, or layout tricks can trigger silent actions that reveal your personal information.
4) Keep up to date with security updates.
Stay up to date with updates from OpenAI, Google, Microsoft and other platforms. Security patches close newly discovered vulnerabilities before hackers can exploit them. Turn on automatic updates to stay protected without having to think about it.
5) Use strong antivirus software.
A strong antivirus program adds another wall of protection. These tools detect phishing links, hidden scripts, and AI-driven exploits before they cause harm. Schedule regular checks and keep your protection up to date.
The best way to protect yourself from malicious links that install malware and potentially access your personal information is to install powerful antivirus software on all your devices. This protection can also alert you to phishing emails and ransomware, keeping your personal information and digital assets safe.
Get my picks for 2025's top antivirus protection winners for your Windows, Mac, Android, and iOS devices at Cyberguy.com.
6) Use multi-layered protection
Think of your security as an onion; more layers make it more difficult to crack. Keep your browser, operating system, and software up to date to ensure endpoint security. Add real-time threat detection and email filtering to block malicious content before it reaches your inbox.
Kurt's key takeaways
AI is advancing faster than most security systems can keep up with. Even as companies quickly patch vulnerabilities, smart attackers are finding new ways to exploit integration and context memory. Staying alert and limiting access to your AI agents is your best defense.
Would you still trust an AI assistant to access your personal email after learning how easily it can be tricked? Let us know by writing to us at Cyberguy.com..
Subscribe to my FREE CyberGuy Report
Get my best tech tips, breaking security alerts, and exclusive offers straight to your inbox. Plus, you'll get instant access to my Ultimate Scam Survival Guide – free when you join my CYBERGUY.COM newsletter.
Copyright CyberGuy.com 2025. All rights reserved.
