MicLinks Radar / Topic

Cybersecurity

Security research, digital threats and practical protection.

Latest in Cybersecurity

5 articles
Hackers Exploit Remote Support Zero-Day
Cybersecurity

Remote-support vulnerability reportedly exploited before disclosure

A critical security vulnerability affecting BeyondTrust Remote Support was exploited by suspected China-linked hackers before it was publicly disclosed. The flaw, identified as CVE-2026-102764, could allow attackers to compromise remote support systems and access sensitive information. Security patches are now available, and organizations running affected systems are urged to update immediately.

Cybersecurity

Microsoft reports active exploitation of a Windows security flaw

Microsoft has released an urgent security update after confirming that attackers are exploiting a Windows vulnerability known as CVE-2026-96951. The flaw can allow malicious files to bypass important Windows security warnings, putting users at risk. Windows users and businesses are advised to install the latest security updates.

Cybersecurity

Reported US inquiry into autonomous AI agents

The US Federal Trade Commission has opened an industry-wide investigation into OpenAI, Anthropic and other AI organisations over potential risks from increasingly autonomous AI agents. The investigation follows security incidents that have raised a difficult question: who is responsible when an AI agent goes beyond the boundaries of the task it was given?

Cybersecurity

Zambia cybersecurity conference puts AI-related threats on the agenda

Cybersecurity professionals from Zambia and across Africa will gather in Livingstone from 28 September to 2 October for the 2026 ISACA Lusaka Chapter Cybersecurity Conference. AI-driven cyberattacks and defence, digital banking security, Zero Trust, cyber resilience and Africa's exposure to cyber warfare are among the major issues on the programme.

Cybersecurity

Local cybersecurity AI: vulnerability research without cloud code uploads

Aikido Security has released Altar, an open-weight AI model designed to run inside an organisation's own infrastructure. It can help analyse security vulnerabilities without sending sensitive source code and internal security information to an external AI provider