Cybersecurity News in Asia

RECENT STORIES:

SEGA moves faster with flow-based network monitoring
Lessons from the Binance hack
Largest cryptocurrency theft on record to date reaches US$6.3bn in sca...
Hikvision showcases its latest range of AI products and solutions desi...
Misleading social media ads promote apps falsely claiming covert surve...
How social-engineered approvals are defeating hardware wallets
LOGIN REGISTER
CybersecAsia
  • Features
    • Featured

      Lessons from the Binance hack

      Lessons from the Binance hack

      Tuesday, November 3, 2026, 1:20 PM Asia/Singapore | Features, Opinions
    • Featured

      How social-engineered approvals are defeating hardware wallets

      How social-engineered approvals are defeating hardware wallets

      Thursday, October 29, 2026, 11:05 AM Asia/Singapore | Features
    • Featured

      Communications infrastructure – from legacy risk to modern resilience

      Communications infrastructure – from legacy risk to modern resilience

      Thursday, October 15, 2026, 2:08 PM Asia/Singapore | Features
  • Opinions
  • Tips
  • Whitepapers
  • AWARDS 2026
  • Directory
  • E-Learning

Select Page

News

Frontier AI firm open-sources coding assistant after silent workspace upload controversy

By CybersecAsia editors | Wednesday, September 23, 2026, 11:20 AM Asia/Singapore

Frontier AI firm open-sources coding assistant after silent workspace upload controversy

Coding assistant silently encrypts and attempts to upload 42,411-file workspace to cloud storage without consent, raising serious doubts about AI agents

After discovering that AI agent could silently package an entire local workspace and attempt to upload it to Alibaba Cloud without clear user consent, Z.AI has open-sourced its coding assistant under the Apache 2.0 license.

The firm has apologized, saying it has removed the feature responsible, and has invited external review of the newly released code, according to Runtime Wire.

The issue had surfaced after an independent developer using the name ferstar examined ZCode while clearing disk space. He found a 313MB encrypted archive representing a commercial workspace containing 42,411 files, along with metadata recording 564 unsuccessful upload attempts.

The collected material reportedly included extensive Git data: the `.git` directory made up 86.6% of the archive, while the Git LFS cache alone accounted for 196.1MB.

According to the reverse-engineering findings, ZCode had obtained upload credentials and an RSA public key from Z.AI’s servers, then encrypted the archive with AES-256-CTR, and then tried to send it to Alibaba Cloud Object Storage.

Because Z.AI controlled the corresponding private key, the user could not decrypt the locally stored archive, raising concerns about what information the company could access if an upload succeeded, according to an AI Weekly report.

What caused the chaos

Z.AI attributed the conduct to a Codebase Indexing function that was enabled by default soon after launch. The feature could create full repository snapshots for a cloud-based Repo Wiki capability.

The firm that in ZCode 3.14.0 it has eliminated Repo Wiki and disconnected the local snapshot-creation and upload paths. It has also said it will establish a vulnerability-reporting program with rewards based on severity and plans to add a no-data-retention option for normal model calls on its platform.

Z.AI has also cited audits by the China Academy of Information and Communications Technology and NSFOCUS, which it says found that the Alibaba Cloud storage bucket contained no remaining data and that relevant stored objects had been permanently erased.

Still, independent observers cannot confirm whether any successfully uploaded files had been previously accessed or used for model training, since Z.AI held the decryption keys.

The dispute has also drawn legal pressure. According to BigGo Finance, Taiyuan Chengming Technology has demanded a written response by 10 October 2026, including a complete account of data processing, deletion evidence, access logs, information on possible third-party sharing or model-training use, and clarification of whether data crossed borders.

Share:

PreviousConsecutive series of agentic AI security breaches expose industry readiness gaps: research group
NextChinese-speaking hacker exploits open source AI agents to hack 100 online retailers

Related Posts

What were the key drivers behind increased 2023 adoption of Zero Trust architecture?

What were the key drivers behind increased 2023 adoption of Zero Trust architecture?

Friday, June 14, 2024

Leaked memo reveals AI firm’s research focus on “rogue“ or “scheming” AI models

Leaked memo reveals AI firm’s research focus on “rogue“ or “scheming” AI models

Friday, February 27, 2026

What three business risks hogged the minds of APAC and global business leaders?

What three business risks hogged the minds of APAC and global business leaders?

Thursday, November 23, 2023

Developing a standards-based DLT security framework for the financial services

Developing a standards-based DLT security framework for the financial services

Thursday, May 21, 2020

Leave a reply Cancel reply

You must be logged in to post a comment.

Voters-draw/RCA-Sponsors

Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
Slide
previous arrow
next arrow

CybersecAsia Voting Placement

Gamification listing or Participate Now

LEARN MORE

Vote Now -Placement(Google Ads)

Top-Sidebar-banner

Whitepapers

  • Critical Security Threatsand the Need for ZTNA: How evolving cyberattacks demand a Zero Trust approach

    Critical Security Threatsand the Need for ZTNA: How evolving cyberattacks demand a Zero Trust approach

    Cyber threats have become more frequent and sophisticated, targeting organizations of all sizes across all …Download Whitepaper
  • Zero Trust Made Simple: Why it matters and how to get started

    Zero Trust Made Simple: Why it matters and how to get started

    Data breaches and cyberattacks are no longer limited to large, high-profile organizations.Download Whitepaper
  • Cloud Secure Edge: Remote access, better security

    Cloud Secure Edge: Remote access, better security

    ​SonicWall Cloud Secure Edge™ is a modern, cloud-native Security Service Edge (SSE) solution that addresses …Download Whitepaper
  • Closing the Gap in Email Security:How To Stop The 7 Most SinisterAI-Powered Phishing Threats

    Closing the Gap in Email Security:How To Stop The 7 Most SinisterAI-Powered Phishing Threats

    Insider threats continue to be a major cybersecurity risk in 2024. Explore more insights on …Download Whitepaper

Middle-sidebar-banner

Case Studies

  • FOMO Pay expands cross-border compliance infrastructure amid rising fraud risks

    FOMO Pay expands cross-border compliance infrastructure amid rising fraud risks

    To unify onboarding and ongoing reviews the payment firm adds identity, liveness, business-ownership, sanctions screening …Read more
  • How a Vietnamese D2C retailer built its own secure digital infrastructure

    How a Vietnamese D2C retailer built its own secure digital infrastructure

    Would your organization build your own digital infrastructure – including AI governance and cybersecurity – …Read more
  • Cyber protection for medical clinics in Singapore

    Cyber protection for medical clinics in Singapore

    As Singapore’s healthcare sector becomes increasingly digital and interconnected, clinics are facing heightened cyber risks, …Read more
  • India’s WazirX strengthens governance and digital asset security

    India’s WazirX strengthens governance and digital asset security

    Revamping its custody infrastructure using multi‑party computation tools has improved operational resilience and institutional‑grade safeguardsRead more

Bottom sidebar

Other News

  • Hikvision showcases its latest range of AI products and solutions designed to promote growth for SMBs

    Friday, October 30, 2026
    HANGZHOU, China, Oct. 30, 2026 …Read More »
  • Cohesity Data Cloud Becomes the First Cloud Recovery Method Validated for Interoperability with IBM Terraform Self-Managed

    Tuesday, October 27, 2026
    Organizations can now rapidly rebuild …Read More »
  • IQuest Research Releases IQuest-Q1 Technical Report, Detailing Training from Scratch and Results Across 30 Benchmarks

    Friday, October 23, 2026
    BEIJING, Oct. 23, 2026 /PRNewswire/ …Read More »
  • New GASA Research: 71% of Adults Surveyed Across 13 Asian Markets Encountered Scams, with Estimated Losses of US$136.5 Billion

    Thursday, October 22, 2026
    Research highlights a gap between …Read More »
  • NEXCOM NGFW platforms secure growing traffic across modern academic networks

    Thursday, October 22, 2026
    TAIPEI, Oct. 22, 2026 /PRNewswire/ …Read More »
  • Our Brands
  • DigiconAsia
  • MartechAsia
  • Home
  • About Us
  • Contact Us
  • Sitemap
  • Privacy & Cookies
  • Terms of Use
  • Advertising & Reprint Policy
  • Media Kit
  • Subscribe
  • Manage Subscriptions
  • Newsletter

Copyright © 2026 CybersecAsia All Rights Reserved.