Sponsored
TEAM BEHIND PROTON MAIL
_Need a password manager you can trust?_
Parse notes
Sponsored
_Need a password manager you can trust?_
and security trusted by 100M+ users and 50,000+ businesses.
Catwatchful, an Android spyware app marketed for "stealthy monitoring," leaked sensitive data from 62,000 users due to an SQL injection vulnerability. The breach exposed email addresses, plain-text passwords, and monitoring data.
TalentHook, a cloud-based applicant tracking system, accidentally exposed the resumes of nearly 26M applicants. Cybernews found an open Azure Blob Storage container with files. The exposed data includes names, emails, addresses, education, and employment history.
Ingram Micro, a business-to-business technology distributor, has been suffering an outage since Thursday. The outage was caused by the SafePay ransomware. The threat actors breached the company through its GlobalProtect VPN.
This write-up explains how to create a "schizophrenic ZIP file". This type of ZIP file appears differently depending on the parser used. The trick relies on how parsers locate the Central Directory, with some using the offset field and others calculating it from the size. As a result, each method points to different Central Directories containing different file content, allowing the same ZIP file to display different contents depending on which parser opens it. This enables malicious files to be opened without the user's knowledge.
Studies have shown that code generated by LLMs contains many security vulnerabilities. Developers should move security earlier in the development process by creating rules files for IDEs like Cursor, which include security guardrails and instructions. This post offers guidance on writing prompts for these rules files. Early research indicates that simply adding the word “secure” to a prompt can significantly reduce vulnerabilities.
TradeTraitor, a North Korean threat actor, has stolen over $2.4 billion through social engineering and cloud attacks, compromising developer credentials to inject malicious code and reroute ETH. Its attacks involve harvesting AWS tokens, injecting JavaScript into cloud applications, with defenses including short-lived tokens, full CloudTrail logs, MFA, and credential monitoring.
Sponsored
Join 1Password on July 15th at 10am PT / 1PM ET [11] for a webinar exploring how teams can take a proactive approach to managing the sprawl of unapproved SaaS apps and AI tools.
Compyl provides solutions for compliance, asset management, contract management, policy management, user access reviews, and risk management (including vendor risk).
A tool to audit Erlang and Elixir dependencies to make sure your Gleam projects sparkle.
LiteLLM is a Python SDK, LLM gateway, and proxy server to call 100+ LLM APIs in OpenAPI format.
Cybersecurity experts report a 19-fold increase in malicious campaigns using Spain's .es domains, making it the third most abused after .com and .ru. Since January, 1,373 subdomains on 447 .es domains have hosted malicious content, with 99% of it focused on credential phishing, targeting Microsoft users via fake HR-themed emails.
A California court has ordered Google to pay $314 million for illegally using Android users' cellular data without consent. The lawsuit revealed that Google transferred device information even when phones were idle, which used up users' data allowances. Google intends to appeal this ruling.
While testing a personal tool, the author of this post noticed that certificates for instagram.com had unusual expiration times. The author set up a machine that attempted to pull the certificates for instagram.com and www.instagram.com every 5 minutes. They discovered that Instagram uses certificates with a lifetime of 8 days, replaces them daily, and uses unique certificates for instagram.com and www.instagram.com.
A China-linked hacking group called Houken exploited three zero-day vulnerabilities in Ivanti Cloud Service Appliance devices to breach French government, telecom, media, finance, and transport organizations starting in September 2024, using sophisticated tools and rootkits to steal credentials and maintain persistent access.
The Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) warns the public about scammers pretending to be federal agents demanding Apple gift cards of $500 or $1,000 to "clear red flags" from accounts, stressing that real federal agencies never ask for payments via gift cards, wire transfers, or cryptocurrency.
IdeaLab, a California tech incubator, suffered a data breach in October 2024 when hackers stole sensitive information.