Hackers are stealing the keys and walking through the front door, and AI is helping them turn the handle (www.techradar.com)

Microsoft’s sixth Digital Defense Report warns that attackers are increasingly using AI to scale credential theft and data extortion, while defenders also adopt AI for detection and response. Over 80% of incidents Microsoft investigated sought data — accessed, exfiltrated, encrypted and ransomed — and humans remain the weakest link: AI-generated, multilingual phishing and social-engineering campaigns make credential harvesting more efficient. Technical takeaways include the rise of infostealing malware (which becomes far less useful when multi-factor authentication is in place), AI-enabled mutating/polymorphic malware that evades signature-based tools, and defenders’ reliance on ML for phishing detection, threat hunting and training. The report also documents an escalation in nation-state activity (notably China, Iran, Russia and North Korea) targeting telecoms, maritime commerce, Ukraine-support networks and commercial IP, often exploiting underfunded or legacy systems. Microsoft calls for wider intelligence sharing, stronger security governance, and reduced ransom incentives as societal measures to deter adversaries. For the AI/ML community this means an accelerating arms race: invest in robust authentication (MFA), adversarial-resilient detection models, behavioral analytics that don’t rely on static signatures, and better collaboration between industry and government to harden critical infrastructure.