The analysis of the 2026 Global Threat Report clearly shows that threats are centered around several strategic areas that defined adversaries’ effectiveness in 2025 – and which will be crucial for building organizational resilience in 2026.
- Adversary Operational Speed
The reduction of the average eCrime breakout time to just 29 minutes (that is 65% faster than in 2024). The fastest recorded attack was just 27 seconds. Reaction time has become a critical factor. Attacks develop faster than traditional decision-making processes in SOCs, and organizations without real-time response automation and correlation operate significantly delayed.
Simultaneously, the report shows that the scale of adversaries is growing – CrowdStrike currently tracks 281 active adversary groups, with 24 new groups identified in 2025 alone. The threat ecosystem is not only accelerating but also expanding.
- Malware-Free Intrusions and Identity Abuse
82% of detections in 2025 concerned activities without the use of classic malware (while in 2020 it was 51%). Techniques based on:
- compromised credentials,
- abuse of legitimate administrative tools,
- manipulating access policies,
- operations in SaaS and the cloud using valid authentication mechanisms
Identity has become the main battleground, and the “identity-first security” model is becoming a foundation.
- AI as a Multiplier of Attack Effectiveness
The year 2025 saw an 89% increase in attacks carried out using artificial intelligence. AI is used for:
- generating credible phishing and vishing campaigns,
- automating reconnaissance,
- creating and modifying code,
- supporting post-exploitation activities.
CrowdStrike Services and CrowdStrike OverWatch responded in over 90 client cases where systems executed malicious code from an attacker utilizing local AI tools (e.g., Claude and Gemini) to generate commands for stealing credentials and cryptocurrencies. AI today plays a dual role: accelerating attackers’ activities and creating a new attack surface. Without appropriate monitoring, AI becomes an uncontrolled risk vector.
- Cloud Under Pressure – Especially in State Operations
35% of cloud incidents were associated with the abuse of valid accounts (valid account abuse). The report points to a 37% increase in targeted attacks on cloud environments year over year, with state-associated cybercriminal groups experiencing an increase of up to 266%. This huge disparity is a clear signal that state operations are focusing on:
- identity environments,
- federations and tenant relationships,
- data in SaaS,
- strategic infrastructure sectors.
- China-nexus and the Logistics Sector
Activity of entities linked to China increased by 38% in 2025, and in the logistics sector by as much as 85%. Of the vulnerabilities they exploited, 40% concerned edge devices exposed to the internet. Attackers consistently targeted, among others, VPN devices, firewalls, gateways, and other perimeter systems, treating these as the preferred vector for initial access.
- Utilization of Zero-day Vulnerabilities and the Shrinking “Security Window”
The report points to a 42% increase in the use of zero-day vulnerabilities before their public disclosure, which shortened the time between the discovery of a flaw and its active exploitation. In practice, this means that organizations may find themselves in a situation where an exploit is operating in a production environment even before the official patch is implemented.
Therefore, not only is the reaction time shortened, but the “security window” between discovering a vulnerability and its operational use by an adversary is also shrinking. In this reality, patch management must be supplemented with segmentation, privilege restriction, and advanced detection, which allow mitigating the attack’s impact even when the vulnerability is already being actively used.