TechBRIEF

Google Launches Selfie Video Sign-In in Taiwan to Help Users Recover Locked Accounts

E
EffectStory 編輯部Editorial Team
Published · Updated
Google introduced a 'Selfie Video' sign-in option in Taiwan this week, letting users regain access to locked accounts by recording a short video of head movements instead of using a trusted device. According to CNA and Inside, the rollout date was July 24, while iThome reported July 23. Google says the videos are encrypted at rest and can be checked at g.co/signin-selfie.

Background and timing of the Taiwan rollout

Google announced the launch of its "Selfie Video" sign-in feature in Taiwan this week, positioned as a way for users to regain access when their account is locked or when they don't have their usual device on hand. According to a report by CNA (中央社), "Google today announced the launch of the 'Selfie Video' sign-in feature in Taiwan" 1, dated July 24, 2026. Inside separately reported the same July 24 launch date, describing it as a new "Selfie video" verification feature for Google Account holders 5.

However, according to a report by iThome, the feature actually rolled out "on Thursday (7/23)," letting eligible users check and enable it from that day 8. The two Taiwan-based dates — July 24 per CNA and Inside versus July 23 (Thursday) per iThome — do not fully align, though all three reports describe the same feature launching in the same week.

OutletReported launch date
CNA2026-07-24
Inside2026-07-24
iThome2026-07-23 (Thursday)

How users set up and use selfie video sign-in

Setup is described as a short, guided process. According to CNA, users "only need to look at the device's camera and follow on-screen prompts to complete a few simple head-turning motions, letting the system capture different angles of their face" 2. Availability is not universal: according to Inside, "eligibility may vary by account," and users can go to g.co/signin-selfie to check whether their account supports the feature and to complete setup 7.

Anti-deepfake and anti-impersonation safeguards

Both CNA and iThome describe layered checks aimed at stopping impersonation. According to CNA, when a user signs in with a selfie video, "Google compares the video captured at that moment against the previously stored selfie video, and requires the user to perform a simple motion to prove this is a live, real-time recording" 3. According to iThome, this design exists specifically "to prevent hackers from using photos, pre-recorded videos, or AI deepfake content to impersonate an account holder"; beyond comparing facial features across the two video clips, the system also "requires users to complete a specific motion as instructed, to prove there is a real person in front of the camera" 9.

How selfie videos are recorded, stored, and the controls users have

Recording only happens after consent. According to CNA, the selfie video is recorded once a user agrees, and the same live-recording proof described above 3 is what the system checks at sign-in time to confirm authenticity. On storage, CNA reports that "user selfie videos are protected by encryption at rest (encrypted at rest), meaning that even when not in use, the video still receives secure storage protection" 4. Users can delete the video at any time, and unless they consent to another use, CNA reports the video is used only to help with sign-in 4.

Where selfie video fits among Google's account recovery options

Selfie video is not a standalone tool but an addition to an existing set of recovery paths. According to Inside, "selfie video is a new account sign-in and recovery option Google has added, following passkeys and backup contacts" 7. According to iThome, Google "currently already offers recovery phone number, recovery email, and recovery contacts as account-recovery methods," and the addition of selfie video means that "even if a user cannot use their original phone, computer, or other verification channels, they can still prove their identity through facial comparison, increasing the chance of recovering the account" 10.

Transparency gaps on retention and matching technology

While Google detailed the consent, encryption-at-rest, and live-motion checks noted above (E3, E4), it has not disclosed everything about how the system works. According to Inside, "Google did not specify the video retention period, nor the detailed technology used to match facial data" 6. This leaves a gap alongside iThome's account of the matching process: iThome describes the system as comparing facial features across two video clips and requiring a proof-of-life motion 9, but neither report explains how long those videos are kept before deletion or what technical method performs the facial comparison.

What this means

Across the three outlets, the core facts agree: Google added a selfie-video sign-in and recovery option in Taiwan this week, built on consent-based recording, at-rest encryption, and a live-motion check designed to block photos, pre-recorded clips, and deepfakes (E1, E2, E3, E4, E5, E9). Yet a factual gap exists on exactly when it launched — July 24 per CNA and Inside versus July 23 (Thursday) per iThome — and a disclosure gap remains on how long the videos are retained and what technology performs the facial matching, a point Inside explicitly flagged as unaddressed by Google 6.

📊 Evidence

FAQ

How can users check if their Google Account supports selfie video sign-in?

According to Inside, eligibility may vary by account, and users can go to g.co/signin-selfie to check whether their account supports the feature and complete setup (E7).

How are selfie videos protected once they're stored?

According to CNA, user selfie videos are protected by encryption at rest, meaning they remain securely stored even when not in use, and users can delete them at any time (E4).

📎 Sources

  1. cna.com.tw
  2. inside.com.tw
  3. ithome.com.tw

Related data

E
EffectStory 編輯部Editorial Team

Related

BRIEF

CNA Launches Taiwan's First News MCP Tool, AskCNA, Priced at NT$200 a Month

Central News Agency (中央社) launched CNA MCP on August 31, 2026, Taiwan's first news tool built on Anthropic's Model Context Protocol (released November 2024), letting AI agents such as Claude, ChatGPT, and Grok retrieve and cite its archives in real time. The tool integrates nearly 5 million newswire stories, 3.5 million photos, and open data from about 150 government agencies, priced at NT$200 a month with an early-bird bonus-quota plan, and received funding from Google Taiwan's nDX Digital Innovation Grant Program.

EffectStory 編輯部 ·
BRIEF

Sony Music and Warner Chappell Sue Anthropic Over Alleged 'Brazen Campaign' of Copyright Theft

Sony Music Publishing and Warner Chappell, joined by other music publishers, sued Anthropic and co-founders Dario Amodei and Benjamin Mann in the U.S. District Court for the Northern District of California, alleging illegal torrenting, scraping, and downloading of copyrighted lyrics and sheet music. The publishers seek up to $150,000 per work and $25,000 per instance of stripped copyright data, a total that could reach several billion dollars. The filing follows Anthropic's earlier $1.5 billion settlement in the Bartz case.

EffectStory 編輯部 ·
BRIEF

Why Anthropic Turned to Nscale and Lambda for $45B and $35B GPU Compute Deals

Anthropic has assembled compute capacity across at least four NVIDIA-linked providers: a $35 billion contract with Lambda tied to a Hut 8-built Texas data center, a $45 billion, six-year deal with Nscale for a West Virginia campus running NVIDIA Vera Rubin systems, a $10 billion contract with startup Volta in Norway, and a reported (unconfirmed) tenancy at Riot Platforms' Rockdale, Texas site. NVIDIA sits inside nearly every arrangement — as investor, lessor, or chip supplier.

EffectStory 編輯部 ·