Back to Articles

Trends

ENISA Threat Landscape 2026: What It Means for NIS2 Entities

12 min read

ENISA's Threat Landscape 2026 analysed 8,257 incidents from 2025: DDoS is 51% of the noise, ransomware does the damage, 60% of known intrusion vectors were vulnerabilities, and 73% of recorded events hit NIS2 entities. What a European company should take from it.

enisa
threat-landscape

ENISA published its Threat Landscape 2026 on 22 September 2026. It is the European Union's reference picture of who attacked whom in 2025, how and why, and it reads differently from vendor reports because it is built on a curated dataset of European incidents. This article goes through the findings that matter to a mid-sized Italian or Spanish company and to anyone in scope of NIS2. It belongs to our series on the year's reports, alongside the cybersecurity trends for 2026, the Verizon DBIR 2026 and the Clusit 2026 report.

Key Takeaways

  • 8,257 incidents from 1 January to 31 December 2025 were analysed, mainly from open sources and anonymised data shared by Member States.
  • DDoS is 51.3% of recorded incidents, largely low-impact hacktivism; ransomware and other financially motivated activity remain the most impactful threat in the short term.
  • Where the intrusion vector was known, 60.4% of unauthorised access incidents leveraged a vulnerability.
  • More than 48,000 CVEs were published in 2025, 22% more than the year before.
  • Essential and important entities account for 73% of recorded events. NIS2 targets the right organisations.

Method and scope

According to the report, ENISA analysts collected and analysed 8,257 incidents, mainly from open sources, plus anonymised information shared by Member States and members of the ENISA Cyber Partnership Programme. This edition moves to the calendar year (1 January to 31 December 2025), so it overlaps by six months with the 2025 edition.

ENISA is explicit about the limits. Open sources do not give a complete picture: ransomware and DDoS are "visible" because their operators claim them, while cyberespionage is often documented months or years later. More reporting about a threat does not necessarily mean more activity. Read the percentages as shares of what was recorded, not as the probability that your company will be hit.

DDoS is the noise, ransomware is the damage

DDoS attacks were the most frequent incident type at 51.3%, followed by unauthorised access at 39.5%. Most DDoS came from ideology-driven hacktivist campaigns timed to elections, protests and geopolitical events, and ENISA notes that these claims "did not result in large-scale or significant impact". ENISA's own NIS investments survey found that organisations in highly critical sectors describe DDoS as "noise", while ransomware dominates their concerns.

By motivation, ideology drove 57.3% of activity and financial gain about 29%. Financially motivated activity, ransomware above all, remained the most impactful threat in the short term. The most active ransomware operations in the EU were Qilin, SafePay, Akira, INC Ransom and Hunters International.

Member StateShare of ransomware claims in the EU (2025)
Germany26.5%
France14.7%
Italy13.6%
Spain12.2%
Netherlands4.7%

Italy and Spain together account for roughly a quarter of the ransomware claims ENISA identified, and the report notes that Qilin's claims targeted mainly France, Spain and Italy.

How attackers got in: vulnerabilities first

The most important number for a technical team is this one: across unauthorised access incidents where ENISA could identify the intrusion vector (5.2% of them), 60.4% leveraged a vulnerability, N-day or zero-day. Misconfiguration and accidental exposure accounted for 20.7%. The sample is small, so treat it as a direction rather than a precise figure, but it points the same way as the Verizon DBIR 2026, which found exploitation overtaking stolen credentials as an entry point.

The vulnerability chapter adds the context:

  • More than 48,000 new CVEs were published in 2025, a 22% increase.
  • 9% were rated Critical and 30% High; 71% listed Network as the attack vector, meaning they can be exploited remotely.
  • The CISA Known Exploited Vulnerabilities catalog added 245 entries during the year.
  • The most common weaknesses among known-exploited vulnerabilities were OS command injection, path traversal and deserialisation of untrusted data.
  • Attackers kept weaponising newly disclosed vulnerabilities "within a constantly decreasing time span", with confirmed attacks on enterprise infrastructure, edge devices and developer tooling.

No team patches 48,000 CVEs. What the numbers argue for is prioritisation by exploitation and exposure: know what you expose to the internet, and fix first what is known to be exploited there. We cover that method in our article on prioritising with KEV, EPSS and SSVC.

Social engineering: phishing kits, AiTM and device codes

Among social engineering techniques, phishing accounted for 77.8% and malicious spam for 13%. Three trends stand out:

  • Phishing-as-a-Service. Tycoon 2FA was reportedly the most widely observed phishing kit globally in 2025. It works as an adversary-in-the-middle (AiTM) proxy designed to bypass multi-factor authentication.
  • ClickFix. Victims are persuaded to "fix" a fake error by running a command themselves, which executes malicious code.
  • Device code phishing and messaging apps. Multiple intrusion sets used trusted platforms such as Signal or WhatsApp to make contact, then steered targets toward legitimate authentication workflows to gain access.

The common thread, in ENISA's words, is a shift "towards techniques designed to further capitalise on abusing trust rather than technical weaknesses alone". For Microsoft 365 and Google Workspace tenants, that makes phishing-resistant MFA and conditional access more than good practice; see our guide to hardening Microsoft 365 and Google Workspace.

Dependencies: one compromise, many victims

The press release headline is about dependencies: "cyber dependencies expand the attack surface". Cybercriminals increasingly targeted third-party providers, such as digital service providers, as an efficient way to reach many victims. Software supply chain compromises, including popular npm packages in the Shai-Hulud campaign, were "of particular concern", and ENISA published a technical advisory on the secure use of package managers in March 2026.

For a NIS2 entity this is Article 21(2)(d) in practice: supply chain security is not a questionnaire sent once a year, it is knowing which suppliers can reach your systems and what happens when one of them is compromised. Our article on NIS2 supply chain security goes into the controls.

Hacktivism, espionage and fraud

Three other threat categories complete the picture:

  • Hacktivism. According to ENISA's announcement, hacktivist claims against EU Member States totalled 4,709, and more than 89% involved DDoS. The report also records more intrusion claims by hacktivist groups in the second half of 2025, notably against operational technology (OT). Most such claims are exaggerated, but an exposed HMI or PLC turns a claim into a real incident.
  • State-nexus activity. Cyberespionage represents a small share of incidents but the most significant strategic threat in the medium to long term. Russia-nexus intrusion sets primarily targeted central government and diplomatic entities, while China-nexus sets showed continuous interest in transport, including maritime. ENISA also notes state-nexus actors using criminal tools, such as Qilin ransomware, which blurs attribution.
  • Fraud. The press release cites a European Banking Authority estimate of about EUR 4 billion lost to online investment fraud across the EEA in 2024, and the report highlights Italy among the countries most affected by fraud and impersonation activity.

For a company, the lesson is less about who the attacker is and more about the entry points they share: ENISA stresses that cybercriminal, hacktivist and state-nexus groups "increasingly rely on similar access vectors, tools and operational approaches".

Sectors, and why NIS2 targets the right ones

SectorShare of recorded incidents
Public administration31.8%
Business services8.5%
Transport8.0%
Manufacturing6.9%
Finance and banking5.6%

Public administration tops the list largely because of hacktivist DDoS. While the dataset includes sectors outside NIS2, essential and important entities represent 73% of the recorded events, which ENISA reads as confirmation of the NIS2 approach.

Manufacturing deserves a closer look for Italian readers. It was hit mostly by unauthorised access (80.4%), not DDoS. Financially motivated incidents made up 61.1% of the sector's total, and ransomware was 82.8% of those, with Qilin, Akira and SafePay the most deployed strains. ENISA names machinery and equipment, construction, automotive, electronics and electrical-equipment manufacturers as particularly affected.

Maturity gaps

The press release also cross-references ENISA's NIS360 assessment: several targeted sectors combine high criticality with lower-than-average maturity, including health, railway, maritime, ICT service management, space, public administration and drinking and wastewater.

AI: speed today, more of the kill chain tomorrow

ENISA observed the continued integration of AI into malicious activity in 2025, for reconnaissance, script development and post-exploitation, and in information manipulation. Its outlook is that AI will "highly likely" increasingly support malicious operations and that 2026 will likely see more phases of the kill chain directly enabled by AI. It also flags the other side: integrating AI into business environments widens the attack surface, and AI systems are themselves targeted.

What a NIS2 entity should take from it

  1. Treat internet-facing vulnerabilities as the first entry point. Keep a live inventory of what you expose, verify which findings are real, and fix known-exploited ones on exposed systems first.
  2. Assume MFA can be phished. Move administrators and high-risk users to phishing-resistant methods and restrict the authentication flows attackers abuse.
  3. Prepare for ransomware, not for headlines. Tested offline backups, segmentation and detection that runs at night and at weekends matter more than DDoS readiness for most mid-sized companies.
  4. Map your dependencies. List the suppliers, managed service providers and software packages that can reach your systems, and decide what you would do if one were compromised.
  5. Measure, then report. NIS2 asks you to assess the effectiveness of your measures. Numbers from your own environment carry more weight with a board than European percentages.

Where Orizon fits

The report's two loudest signals, exploited vulnerabilities and ransomware, map onto two of our services. RECON maps your external, internal and cloud attack surface and verifies each finding, so you know which exposures are real and which known-exploited vulnerabilities sit on systems an attacker can reach. Oversight, our SOC as a Service on the CYREBRO platform, watches your sources 24/7 and delivers each validated investigation with the evidence and the remediation steps, which is where ransomware is stopped or not. For the governance side, our NIS2 compliance page shows how these feed Article 21 evidence.

If you want to see your own numbers next to ENISA's, start with a RECON scan of your domains: you will know in a day what you expose and which findings are proven.

enisa
threat-landscape
ransomware
nis2
vulnerabilities
europe