ENISA, the EU's cybersecurity agency, published its Threat Landscape 2026 on September 22, 2026. It analyses 8,257 incidents recorded between January 1 and December 31, 2025. Denial-of-service attacks made up 51.3% of them and unauthorised access 39.5%. Public administration was the most affected sector at 31.8%. As of November 13, 2026 this is the current edition. The figures below are from the report itself.
Headline Numbers
| Measure | Figure |
|---|---|
| Incidents analysed | 8,257 |
| Period covered | January 1 to December 31, 2025 |
| DDoS share of incidents | 51.3% |
| Unauthorised access share | 39.5% |
| Ideology-driven incidents | 57.3% |
| Financially motivated incidents | 29.3% in the summary, 29.2% in the overview chapter |
| Cyberespionage | 5.9% in the summary, 6.3% in the overview chapter |
| Hacktivist claims against EU member states | 4,709, more than 89% of them DDoS |
| Share of incidents hitting essential entities under NIS2 | 72.9% |
| New vulnerabilities with a CVE identifier in 2025 | More than 48,000, up 22% |
The report gives two slightly different figures for financial motivation and for espionage in different chapters. We show both.
By Sector
| Sector | Share of incidents | Main incident type |
|---|---|---|
| Public administration | 31.8% | DDoS. Ideology-driven DDoS was 81.8% of the sector's incidents |
| Business services | 8.5% | Unauthorised access, 56.7% |
| Transport | 8% | DDoS claims, 83.1% |
| Manufacturing | 6.9% | Unauthorised access, 80.4% |
| Finance and banking | 5.6% | Unauthorised access, 46% |
Manufacturing stands out. Financially motivated attacks were 61.1% of its incidents, and ransomware was 82.8% of those. Across all sectors, manufacturing took 25.2% of ransomware deployments.
How Attackers Got In
- Among social engineering techniques, phishing accounted for 77.8% and malicious spam for 13%.
- ENISA could identify the way in for only 5.2% of unauthorised access incidents. Of those, 60.4% used a software vulnerability and 20.7% a misconfiguration or accidental exposure.
- For state-linked activity, a way in was identified in about 20% of incidents, and 70% of those used a vulnerability.
Ransomware and Cybercrime
| Measure | Figure |
|---|---|
| Ransomware share of financially motivated claims | 47.3% |
| Data breach share | 36% |
| Fraud and impersonation share | 13.3% |
| Germany's share of ransomware claims | 26.5% |
| France | 14.7% |
| Italy | 13.6% |
| Spain | 12.2% |
| Netherlands | 4.7% |
ENISA's press release gives a different split, with ransomware at 40% of analysed cybercrime events, data breaches at 31% and fraud at 19%. The two sets are calculated on different bases, so use the report's figures and name the base.
Hacktivism
One pro-Russia group accounted for 48% of all ideology-driven activity. ENISA stresses that these attacks were mostly low impact and short. It notes a rise in claimed intrusions into operational technology and industrial sites, with 37 such claims in the third quarter of 2025, though verified disruption stayed limited. Germany, France, Italy, Spain and Poland were among the most targeted states. For what that looked like a year later, see our refrigeration attack timeline and operational technology security.
Compared With the 2025 Edition
| Measure | 2025 edition | 2026 edition |
|---|---|---|
| Published | October 1, 2025 | September 22, 2026 |
| Period | July 2024 to June 2025 | Calendar year 2025 |
| Incidents | 4,875 | 8,257 |
| DDoS share | 77% | 51.3% |
| Ideology-driven share | Almost 80% | 57.3% |
| Public administration | 38.2% | 31.8% |
| Transport | 7.5% | 8% |
| Manufacturing | 2.9% | 6.9% |
| Finance | 4.5% | 5.6% |
| Essential entities under NIS2 | 53.7% | 72.9% |
These are not like-for-like. The two periods overlap by six months, and ENISA says it expanded the cybercrime activity it tracks, including data breaches and fraud, which it says affected the numbers without changing the rankings. The fall in the DDoS share is therefore partly a change in what was counted. ENISA also revised some figures in the 2025 edition on September 22, 2026, so the 2025 column, taken from the original press release, may differ slightly from the current PDF.
What ENISA Says About AI
The report finds that attackers in 2025 mainly used consumer AI tools to speed up existing methods, not to achieve new capabilities. Looking ahead, it assesses that AI will highly likely support more malicious operations and that 2026 is likely to see more phases of an attack directly enabled by AI. Cases are collected in documented AI-enabled cyberattacks.
Methodology
All 2026 edition figures were read from the ENISA Threat Landscape 2026 report (PDF) and checked against ENISA's press release of September 22, 2026. The 2025 edition figures are from ENISA's publication page and its press release of October 1, 2025. ENISA's data comes mainly from open sources, plus anonymised information from member states. It warns that more reporting of a threat does not necessarily mean more of that threat. Read November 13, 2026.