Microsoft Outlook Add-In Breach Exposes Thousands of Accounts in Bold Credential Heist
Over 4,000 Microsoft account credentials vanished through a forgotten Outlook add-in after cybercriminals hijacked the AgreeTo project abandoned in December 2022. Attackers claimed the orphaned Vercel-hosted URL, transforming the dormant add-in into a phishing weapon that displayed fake login pages directly within Outlook’s sidebar—exfiltrating passwords, credit card details, and banking credentials via Telegram Bot API during Microsoft’s marketplace continued hosting it for months. Koi Security researchers exposed the “AgreeToSteal” operation, revealing how weak marketplace oversight and absent post-approval monitoring created the perfect storm for credential theft, with victims redirected to legitimate login pages after compromise to mask the breach.
When a developer steps away from a project, cybercriminals are often waiting in the wings to take advantage. That’s precisely what occurred with AgreeTo, an abandoned Outlook add-in that became the perfect vehicle for a massive phishing operation researchers have dubbed “AgreeToSteal.” Over 4,000 Microsoft account credentials were stolen through this supply chain attack, demonstrating once again that neglected software is a boon to threat actors.
Abandoned software creates perfect opportunities for cybercriminals, turning neglected projects into weaponized tools that compromise thousands of unsuspecting users.
The scheme was elegantly simple yet devastatingly effective. After the original developer abandoned the AgreeTo project in December 2022, an unknown attacker claimed the Vercel-hosted URL (outlook-one.vercel.app) that the add-in still referred to. Microsoft’s store continued to list the add-in despite its orphaned status, creating a window of opportunity that remained open for months. The attacker exploited this gap between abandonment and platform oversight, hijacking the entire infrastructure without raising immediate alarms.
Here’s where things get nasty. When victims opened the compromised add-in, a convincing fake Microsoft sign-in page appeared in their Outlook sidebar. Credentials entered into this phishing page were quietly exfiltrated via Telegram Bot API before users were redirected to the legitimate Microsoft login page to avoid suspicion. The phishing kit was comprehensive: fake sign-in interface, password collection mechanisms, exfiltration scripts, and seamless redirects that left victims none the wiser.
But stolen passwords were just the beginning. The attacker harvested credit card numbers, CVVs, PINs, banking security answers, and even Interac e-Transfer credentials from compromised accounts. Koi Security researchers who discovered the breach accessed the attacker’s exfiltration channel and witnessed something chilling—the threat actor actively testing stolen credentials in real time.
The permissions granted by the original manifest created an even scarier potential scenario. With ReadWriteItem permissions intact, the attacker had authorisation to read and modify user emails. Although no covert email exfiltration was confirmed, Koi Security warned that the capability existed to deploy JavaScript for complete mailbox siphoning. That’s access most of us would never knowingly grant a stranger.
Microsoft’s marketplace vulnerabilities facilitated this entire operation. The platform examines manifest files only at submission, with no periodic content monitoring afterward. Once approved, add-ins load resources from developer servers without ongoing verification—essentially distributing a simple URL with minimal security oversight. The attacker bypassed this weak vetting process through straightforward URL hijacking. Office add-ins use a manifest file that fetches content from a developer’s server, creating an architectural vulnerability that Microsoft’s initial review process fails to address for dynamic content changes. This marks the first documented case of malware hosted directly on the Microsoft Marketplace.
Investigation revealed this wasn’t an isolated incident. The same threat actor operates at least a dozen additional phishing kits targeting banks, webmail providers, and internet service providers simultaneously. It’s a coordinated campaign exploiting marketplace infrastructure across multiple platforms.
Microsoft removed the add-in following Koi Security’s announcement, and all 4,000 compromised victims received breach notifications. However, the incident raises uncomfortable questions about software abandonment protocols and marketplace oversight that won’t simply vanish with one takedown.
Final Thoughts
The recent breach involving Microsoft Outlook add-ins underscores a significant vulnerability in enterprise security—third-party extensions can compromise even the most secure systems. While Microsoft’s platform is designed to protect users, it cannot prevent all risks associated with external add-ins. This incident serves as a reminder to rigorously audit your add-ins, implement multi-factor authentication, and critically assess permission requests.
At Computer Repair Geeks, our team is here to help you strengthen your security measures and safeguard your organization against such vulnerabilities. Don’t leave your enterprise exposed; take proactive steps to protect your data.
Ready to enhance your security? Click on our “Contact Us” page to get in touch with our experts today!