Multi-factor authentication, long considered a cornerstone of identity security, is facing a growing threat that bypasses its protections not through technical exploits but by exploiting human psychology. Attackers are increasingly deploying "prompt bombing" campaigns that flood users with authentication requests until fatigue drives them to approve a login they should have denied.
According to The Hacker News, which reported on the issue on 26 May, the attack method sidesteps the fundamental premise of MFA entirely. Rather than attempting to compromise the second factor directly, threat actors simply wait for the legitimate user to surrender it voluntarily through repeated push notification spam.
The mechanics are straightforward: once an attacker obtains valid credentials, they trigger a cascade of MFA approval prompts to the target's mobile device or authenticator app. After dozens of notifications—often arriving in rapid succession and sometimes outside normal working hours—many users approve the request simply to silence the interruptions. The attacker gains access with the user's unwitting cooperation.
Security analysts have long warned that convenience-focused push approval workflows create a predictable failure point. The vulnerability lies not in the cryptographic strength of MFA itself, but in the assumption that users will scrutinize every authentication request with equal vigilance. In practice, cognitive overload and notification fatigue consistently win.
The industry response is shifting toward automated, risk-aware authentication frameworks that reduce reliance on user judgment. Number-matching protocols, which require users to enter a code displayed on the login screen into their authenticator app, have proven effective at eliminating blind approvals. Microsoft and other identity providers have begun making number-matching the default for push-based MFA.
Beyond workflow changes, adaptive authentication engines that evaluate contextual signals—geolocation anomalies, unfamiliar device posture, impossible travel velocities—are gaining traction. These systems can suppress unnecessary prompts or trigger step-up verification only when risk thresholds are crossed, reducing the attack surface for prompt bombing while maintaining security posture.
FIDO2-compliant hardware security keys and platform biometrics offer another layer of protection, as they resist fatigue-based manipulation by design. However, deployment costs and legacy system compatibility remain barriers for many organizations considering a full migration.
For IT teams managing enterprise identity infrastructure, the threat underscores a broader lesson: zero-trust architectures must decouple authentication success from manual user intervention wherever possible. Training programs should emphasize verification of login context—checking timestamps, locations, and device information—rather than conditioning users toward blanket approval habits.
Rate-limiting on authentication requests represents a low-cost, high-impact mitigation that organizations can implement immediately. By capping the number of MFA prompts an account can receive within a given window, security teams can disrupt the flooding mechanism that makes prompt bombing viable in the first place.
As identity becomes the primary perimeter for modern enterprises, the industry is learning that the weakest link in multi-factor authentication is often the human asked to validate it.
2022年，攻擊者透過 MFA prompt bombing 成功入侵 Cisco 系統，竊取 2.8GB 資料，這宗事件突顯了多重身份驗證正面臨的日益嚴重威脅。攻擊者並非透過技術漏洞繞過其保護，而是利用人類心理弱點，發動「prompt bombing」攻擊，向用戶發送大量身份驗證請求，直至用戶因疲勞而批准本應拒絕的登入。
The Hacker News 於5月26日報道此問題，指該攻擊方法完全繞過了MFA的基本前提。威脅行為者並非嘗試直接破解第二因素，而是透過重複的推送通知垃圾訊息，等待合法用戶自願交出驗證權限。
攻擊手法相當直接：攻擊者取得有效憑證後，便會向目標的流動裝置或驗證應用程式發送一連串MFA批准提示。經過數十個通知——通常密集出現，有時甚至在正常工作時間以外——許多用戶只是為了停止騷擾而批准請求。攻擊者便在用戶不知情的情況下獲得存取權限。
安全分析員長期警告，以方便為導向的推送批准工作流程會造成可預見的失誤點。漏洞不在於MFA本身的加密強度，而在於假設用戶會以同等警惕性審查每個身份驗證請求。實際上，認知超載和通知疲勞往往佔上風。
業界應對方案正轉向自動化、具風險意識的身份驗證框架，減少對用戶判斷的依賴。號碼配對協議要求用戶將登入畫面顯示的代碼輸入驗證應用程式，已證明能有效消除盲目批准。Microsoft 及其他身份供應商已開始將號碼配對設為推送式MFA的預設設定。
除了工作流程變更外，自適應身份驗證系統評估環境信號——地理位置異常、不熟悉裝置狀態、不可能的移動速度——正獲得廣泛採用。這些系統可在風險閾值被觸發時才抑制不必要的提示或啟動升級驗證，在維持安全態勢的同時減少 prompt bombing 的攻擊面。
符合 FIDO2 標準的硬件保安金鑰和平台生物識別提供另一層保護，因為它們在設計上便能抵抗基於疲勞的操控。然而，部署成本和舊系統兼容性仍是許多考慮全面遷移的機構面臨的障礙。
對於管理企業身份基礎設施的 IT 團隊而言，此威脅突顯了一個更廣泛的教訓：零信任架構必須盡可能將身份驗證成功與手動用戶干預分離。培訓計劃應強調驗證登入環境——檢查時間戳、位置和裝置資訊——而非訓練用戶養成一律批准的習慣。
對身份驗證請求實施頻率限制是一項低成本、高效益的緩解措施，機構可立即實施。透過限制帳戶在特定時間內可接收的MFA提示數量，安全團隊可破壞使 prompt bombing 可行的泛濫機制。
隨著身份成為現代企業的主要防禦邊界，業界正認識到多重身份驗證中最薄弱的環節往往是負責驗證的人類。