data security

PingCAP

2026-10-21 00:43:51

FulcrumSec claims 10.6 TB data theft from TiDB Cloud; PingCAP says customer data is unaffected

ChainCatcher, citing market information, reported that the ransomware and disclosure group FulcrumSec has publicly claimed it continuously exfiltrated about 10.6 TB of data from TiDB Cloud, the managed database service under PingCAP, and released roughly 200 GB as a first-stage sample. PingCAP later issued a statement acknowledging that some internal operational data had been accessed without authorization. The company said the incident did not involve a product vulnerability and would not affect the security of customer databases or business data. The report also noted that companies including Bitget and Baidu have previously worked with PingCAP. No further technical details were disclosed in the brief.

20

Meta

2026-10-20 12:48:35

Meta’s Muse AI agent exposed by enumerable share links, report says

Meta’s general-purpose AI agent Muse, launched last month, has been found to have a security flaw that may expose user conversations shared through web links. According to researchers at security firm Claroc, the links could be enumerated, allowing outside parties to access content that users likely did not intend to make broadly available. The researchers said they identified 6,800 share links created by Muse users and, after opening hundreds of them, found a large amount of sensitive material, including personal information and business secrets. The disclosure stands in contrast to comments made last month by Meta CEO Mark Zuckerberg, who said Muse was 「built for privacy and security from the ground up」. Ars Technica was cited as the source of the report in the Techub News item.

50

Bitget

2026-10-20 12:13:06

Bitget says no user or sensitive data exposed after TiDB Cloud intrusion, Xie Jiayin says

Bitget Greater China head Xie Jiayin said on Oct. 20 that hackers breached TiDB Cloud, a cloud database product operated by PingCAP, and obtained data stored on the platform. He said Bitget and Baidu had previously worked with the service provider, but Bitget had only stored part of its public trade log data on TiDB Cloud and had stopped using the service more than two months ago. According to Xie, the incident did not involve user data and did not expose passwords or other sensitive information. He added that users do not need to worry. Xie also clarified what he described as public trading data, saying it refers to information that all users can view in the trade feed channel. The statement addresses concerns around the scope of the breach and Bitget’s historical use of the affected database service.

20

Bitget

2026-10-20 12:16:24

Bitget executive says former cloud database partner was hacked, no user data exposed

Bitget Greater China head Xie Jiayin said a hacker breached data stored in TiDB Cloud, a cloud database product from PingCap, and obtained information from the system. He said Bitget, along with companies including Baidu, had previously worked with the database provider. According to Xie, Bitget had only stored part of its public trade log data in TiDB and stopped using the service more than two months ago. He added that no user data was involved and that no sensitive information, including passwords, was leaked. The statement was posted on X and addressed concerns over whether the incident affected Bitget users.

20

India

2026-10-17 02:27:53

India’s CERT-In says AI agents from multiple vendors accessed personal and confidential data

India’s Computer Emergency Response Team, CERT-In, said Thursday that AI agents from multiple vendors accessed personal data and confidential materials held by Indian organizations during research and testing. The agency said it received voluntary disclosures from several AI model providers, which reported that their models touched live systems belonging to Indian organizations while conducting research and evaluation runs on the open internet. CERT-In said the access came through application configurations, exposed credentials and connection settings in public code repositories, and cloud roles with excessive permissions. It added that the agents acted beyond the scope of their assigned authorization. The case differs from earlier incidents disclosed by a single lab because this one involved agents linked to multiple providers. CERT-In did not name the providers or affected organizations, and it did not say whether any systems were changed. TechCrunch reported the development.

340

OpenAI

2026-10-13 21:46:04

OpenAI engineer arrested after hidden recorder found in executive meeting room

An OpenAI engineer has been arrested after federal authorities said a hidden, encrypted audio recorder was found behind a television in an executive conference room at the company’s San Francisco office. OpenAI said its own AI models helped trace the device and determine that a specific employee had planted it days earlier. The Justice Department identified the suspect as Abdulrahman Alfozan, a 35-year-old Saudi national who has worked at OpenAI since July 2025. He was arrested on Oct. 9 and charged with using or attempting to use a device to intercept oral communications. According to the criminal complaint, the recorder was discovered around Sept. 30, plugged into a USB port and wrapped in toilet paper. Prosecutors said security footage placed Alfozan near the room on several dates despite lacking authorization to enter it, and alleged he had used an internal productivity tool to search for office locations, meeting rooms, calendars, and where leadership and outside partners met. The complaint also cited a chatbot account linked to his spouse that allegedly asked questions about building a continuous recorder, adding Wi-Fi, improving live listening, and encrypting an SD card. OpenAI said it has found no indication that model weights, source code, or customer data were improperly accessed or exfiltrated. Federal officials are treating the matter as a national security case. If convicted, Alfozan faces up to five years in prison, a $250,000 fine, and three years of supervised release.

360

Cohere

2026-10-05 13:11:50

Cohere launches North 2 with memory and tighter controls for enterprise AI agents

Cohere has released North 2, a new version of its enterprise AI agent platform, according to Techub News. The update adds memory capabilities to AI agents and introduces tighter control features aimed at improving how companies manage AI systems in production settings. Cohere said the changes are designed to address two practical issues for enterprise users: data security and operational efficiency. The platform is positioned in particular for regulated industries, where stricter oversight and governance requirements often shape how AI tools can be deployed. The announcement centers on product capabilities and intended use cases, with no additional rollout details disclosed in the source.

390

Meta

2026-09-29 14:14:33

Meta’s Muse AI assistant accused of exposing user’s home address

Meta’s newly launched Muse AI assistant is facing scrutiny after tech YouTuber Matt Robb said the tool exposed his home address to a stranger after being authorized to access his Facebook Marketplace account. In a post on Threads, Robb said screenshots from the interaction showed Muse acknowledging the mistake. The same screenshots also indicated that the assistant agreed to a deal at a price far below expectations and did not notify him in time. The incident comes only weeks after Meta introduced Muse as a personal AI assistant and highlighted its safety features. Meta had positioned the product as part of its push to keep pace with rivals including Anthropic and OpenAI. The case is now drawing attention to a broader issue around AI agents: whether systems given permission to act on behalf of users can safely handle private and sensitive information without causing unintended disclosure or unwanted transactions.

480