| Mobile Web

Search results for SecurityWeek

AI & Enterprise

Hackers tried to exploit PHP flaws four days before patch, could be Log4Shell-scale

Hackers began searching for vulnerable servers and attempting attacks four days before patches for two critical PHP flaws were released, SecurityWeek reported on Oct. 22 citing Amazon. The flaws, known as PHP2Shell, can be used to remotely control servers. Amazon said exploit activity accelerated after information leaked online and that PHP2Shell is now widely abused. Researchers and companies estimate millions of PHP hosts remain unpatched, and CISA urged immediate updates.

AI & Enterprise

Meta GitHub repositories hacked, some source code and data on about 1,500 people stolen

Hackers stole Meta employee account credentials and broke into parts of the company’s GitHub repositories, taking product source code and personal data on about 1,500 people, SecurityWeek reported on Oct. 13. Most of the victims were Meta employees in the United States. Meta confirmed the intrusion and notified victims. It said 99 percent of the stolen data was software code and 1 percent was personal information.

AI & Enterprise

Zero trust still valid in AI era, but must be implemented properly

John Kindervag, who created the concept of zero trust, has published a new book arguing the model remains effective in the AI era. Kindervag first introduced zero trust in a 2010 Forrester Research report, and it later became a core cyber security principle. He said experts concluded AI threats are faster, more sophisticated and larger in scale, but fundamentally unchanged. The key, he said, is correct implementation through a policy engine that reflects an organisation’s security posture.