Bouygues Telecom, France’s third-largest mobile operator, has confirmed a significant cyberattack that compromised the personal data of approximately 6.4 million customer accounts.

Researchers have uncovered critical vulnerabilities in end-to-end encryption (E2EE) used in TETRA-based radios, which are widely deployed by police, military, and intelligence agencies globally.

A new clickjacking campaign is exploiting adult content websites hosted on blogspot[.]com to distribute a Trojan that manipulates Facebook interactions. This campaign leverages malicious SVG files containing obfuscated JavaScript.

Air France and KLM have disclosed a data breach resulting from unauthorized access to a third-party customer service platform. The breach exposed limited customer data but did not compromise sensitive information such as passwords or payment details.

Researchers revealed critical vulnerabilities in satellite and ground station software that could allow attackers to hijack or disable space assets. Exploitation of these bugs could allow attackers to take full control of satellite systems and more.

A newly leaked trove of data has exposed the inner workings of North Korea’s covert IT worker operations, revealing how the regime systematically places skilled developers in remote jobs worldwide to generate revenue for its weapons programs.

Context corruption is the most critical threat to agentic AI systems. LLMs are unreliable at distinguishing between legitimate and malicious instructions. Attackers can inject malicious prompts that alter the agent’s behavior.

A sophisticated cybercriminal operation led by the VexTrio group has been uncovered, involving the distribution of fraudulent mobile applications through the Google Play Store and Apple App Store.

Google has confirmed a data breach involving one of its Salesforce instances, attributed to the threat actor group UNC6040, linked to the notorious ShinyHunters. The attackers accessed contact information and related notes of SMB customers.

The vulnerability (CVE-2025-53786) allows attackers to manipulate user credentials, escalate privileges, and impersonate hybrid users. It also exposes Entra ID, potentially compromising cloud-based identity services.