Newsgather
BackGoogle Launches Fake Call Detection for Android to Combat AI Voice Deepfakes
Google Launches Fake Call Detection for Android to Combat AI Voice Deepfakes
En développement
ITmedia03.06.2026Tech2 dk okumaJapan

Google Launches Fake Call Detection for Android to Combat AI Voice Deepfakes

L'essentiel

  • Google announced "fake call detection" for Android to combat AI voice deepfake scams.
  • Rolling out globally on Pixel devices, it verifies genuine calls using RCS, warning users if a call is fraudulent.

Résumé généré par IA

Pourquoi c'est important

Attackers are increasingly using voice cloning technology to impersonate family members or superiors, often faking emergencies to demand money. INTERPOL highlighted these impersonation scams as a major contributor to over $400 billion in global damages.

Taille de police

Google announced on June 2nd (local time) that it will release a new feature called "fake call detection" for Android to prevent "impersonation call scams" using AI-generated voice deepfakes. The company stated that it plans to begin a global rollout starting this month, initially with the Pixel series.

In recent years, attackers have been using tactics such as spoofing phone numbers to appear as if they are calling from a contact person, and combining this with AI voice cloning technology to impersonate family members or superiors, faking emergencies to demand money. Google explained that this initiative is in response to INTERPOL's mention in March of impersonation scams as one of the main causes of over $400 billion in damages worldwide.

Fake call detection automatically operates when both the caller and receiver are using Google's Phone app. The caller's device sends a "silent confirmation signal" to the receiver in real-time, digitally verifying that it is a genuine call. This "digital handshake" uses "RCS", and the communication content cannot be referenced by external parties.

If an attacker impersonates a contact, this confirmation signal will be missing, prompting the receiver's device to reconfirm with the actual contact if they are currently making the call. If the actual device responds that it is "not making a call," a warning will be displayed on the screen instructing the user to "hang up immediately." Google described this as an "industry-first protection feature."

To use the feature, the following conditions must be met:

Devices with Android 12 or later

Google's "Phone," "Contacts," and "Messages" installed

RCS enabled in Google's "Messages"

Both the caller and receiver using Google's "Phone"

While Google's "Phone" is the default phone app on many Android devices, if you use another app, you can install it from the Play Store and set it as the default to use this feature. This feature is enabled by default but can also be disabled in the app's settings.

Google stated that because this feature is built on the open standard RCS, other app and device manufacturers can also adopt it. The company is already advancing its anti-scam measures, such as AI-powered Scam Detection in "Messages" and "Phone" (available on Pixel and Samsung devices), and fake call detection is an extension of these efforts.

Questions ouvertes

  • Will this feature be available on non-Pixel Android devices immediately?
  • What is the exact technical implementation of the 'silent confirmation signal'?
  • How effective is the system against sophisticated deepfake audio?
  • What are the privacy implications of RCS being used for call verification?

Sujets liés

This article was originally published by ITmedia.

Articles liés

ファイル転送サービス「ギガファイル便」で障害発生、代替ドメインで暫定運用へ
En développement·1 sa önce

ファイル転送サービス「ギガファイル便」で障害発生、代替ドメインで暫定運用へ

7月6日、ファイル転送サービス「ギガファイル便」で障害が発生し、多くのユーザーに影響が出た。運営元は代替ドメイン「gigafile.jp」での暫定運用を開始し、従来のURLもドメイン部分を書き換えることでアクセス可能とした。ユーザーは一時的にログアウト状態となり、再ログインが促されている。

ITmedia
BIGLOBEメールと@niftyメールで計726万件超のメールアドレスとID漏洩、パスワードも一部流出
En développement·5 sa önce

BIGLOBEメールと@niftyメールで計726万件超のメールアドレスとID漏洩、パスワードも一部流出

BIGLOBEとニフティは、KDDIのメール基盤への不正アクセスにより、計726万件以上のメールアドレスとBIGLOBE IDが漏洩したと発表。BIGLOBEでは501万件超、ニフティでは224万件超が対象で、一部はパスワードも漏洩。両社はパスワード変更を呼びかけ、未変更者にはリセットや無効化を実施している。

ITmedia
Plus sur ce sujetGoogle