AI-Driven Cyberattack Targets Adif and Renfe: 500GB Data Exfiltrated from Spanish Railway ...

A highly sophisticated cyberattack leveraging artificial intelligence (AI) has targeted the Spanish railway infrastructure operator Adif and the national railway company Renfe , resulting in the exfiltration of approximately 500GB of data. This incident, first reported by El Mundo on 25 September 2026, marks a significant escalation in the threat landscape for critical infrastructure, as it is one of the first publicized cases in Spain where AI was explicitly used to automate the entire attack chain. The attackers utilized AI-driven tools to autonomously discover and exploit vulnerabilities, enabling rapid lateral movement from Adif ’s web infrastructure to its cloud environment and subsequently to Renfe ’s systems. While Renfe asserts that only limited, non-sensitive data was accessed, the attackers are believed to have targeted traveler databases, which are highly valuable on the black market. The breach has prompted an ongoing investigation by the Centro Criptológico Nacional (CCN-CERT) , part of the Spanish National Intelligence Centre ( CNI ), and highlights the urgent need for advanced monitoring, rapid patching, and robust incident response capabilities in the face of evolving AI-driven threats.

Technical Information

The attack on Adif and Renfe demonstrates a new paradigm in cyber threats, where AI is not merely a tool for defenders but is now actively weaponized by adversaries to automate and accelerate the entire cyber kill chain. The following technical analysis provides a comprehensive breakdown of the incident, the tactics, techniques, and procedures (TTPs) employed, and the broader implications for organizations operating critical infrastructure.

Attack Vector and Initial Access

The attackers deployed an AI system, reportedly with capabilities similar to those developed by Anthropic , to conduct large-scale, automated vulnerability scanning and fuzzing against Adif ’s public-facing web infrastructure. The AI autonomously identified a backdoor—likely a previously unknown or unpatched vulnerability—in one of Adif ’s web applications. This vulnerability was exploited to gain initial access, consistent with the MITRE ATT&CK technique T1190: Exploit Public-Facing Application .

The use of AI in this context enabled the attackers to rapidly enumerate exposed services, fingerprint application stacks, and test for a wide array of vulnerabilities at a scale and speed unattainable by manual methods. This approach significantly reduces the window of exposure for unpatched systems and increases the likelihood of successful exploitation, even in environments with robust perimeter defenses.

Lateral Movement and Privilege Escalation

Upon establishing a foothold in Adif ’s web server, the attackers leveraged their access to pivot into the organization’s cloud infrastructure. This lateral movement suggests a high degree of familiarity with Adif ’s internal architecture and may have involved the exploitation of weak inter-system authentication, misconfigured cloud permissions, or the abuse of privileged credentials harvested during the initial compromise.

The attackers then extended their reach to Renfe ’s interconnected web systems, indicating that the two organizations’ environments were either directly linked or shared common authentication and access control mechanisms. The techniques observed align with MITRE ATT&CK tactics such as T1210: Exploitation of Remote Services and T1526: Cloud Service Discovery .

The primary objective of the attack appears to have been the exfiltration of sensitive data, with approximately 500GB stolen over several days. While Renfe has stated that only “limited, non-sensitive” information was accessed—explicitly excluding payment data and national identification numbers—open-source intelligence suggests that the attackers targeted traveler databases. Such data typically includes names, travel histories, information, and potentially other personally identifiable information (PII), all of which are highly valuable for identity theft, social engineering, and black-market resale.

The exfiltration was likely conducted using encrypted channels or covert data transfer techniques to evade detection by traditional security monitoring tools. This aligns with MITRE ATT&CK technique T1567: Exfiltration Over Web Service .

Detection, Response, and Attribution

The breach was detected after several days of sustained activity, with the most significant data theft occurring on Thursday, 24 September 2026. The incident was promptly reported to CCN-CERT , which is now leading the investigation and providing remediation guidance.

Attribution remains unclear, as no criminal group has publicly claimed responsibility, and there are no direct links to known advanced persistent threat (APT) groups. However, the TTPs observed—AI-driven automated exploitation, lateral movement in cloud environments, and targeting of critical infrastructure—are consistent with those used by groups such as TA2541 , Leviathan (APT40) , TeamTNT , and the Play Ransomware Group . It is important to note that none of these groups have been directly linked to this incident as of this report.

Indicators of Compromise (IOCs) and Affected Assets

The confirmed affected domains are adif.es and renfe.com . No specific malware samples, exploit code, or detailed IOCs have been published in open sources at this time. The attack leveraged the following MITRE ATT&CK techniques:

T1190: Exploit Public-Facing Application

T1119: Automated Collection

T1046: Network Service Scanning

T1210: Exploitation of Remote Services

T1526: Cloud Service Discovery

T1567: Exfiltration Over Web Service

No public disclosure of specific product versions, software platforms, or CVEs has been made. The only confirmed affected assets are Adif ’s web servers and cloud infrastructure, and Renfe ’s interconnected web systems.

Exploitation in the Wild and Broader Implications

This incident is notable as one of the first reported cases in Spain where AI was explicitly used to automate the attack chain against public sector infrastructure. It follows a broader trend of increasing attacks on Spanish critical infrastructure and utilities, as evidenced by recent incidents targeting Endesa and Movistar .

The use of AI in offensive cyber operations represents a significant escalation in both the speed and scale of attacks. AI-driven tools can autonomously discover and exploit vulnerabilities, adapt to changing environments, and evade detection through the use of polymorphic techniques and encrypted communications. This raises the bar for defenders, who must now contend with adversaries capable of launching highly automated, adaptive, and persistent attacks.

In light of this incident, organizations operating critical infrastructure should take the following actions:

Conduct a comprehensive forensic review of all web-facing applications to identify evidence of automated scanning and exploitation. Audit cloud infrastructure for unauthorized access, privilege escalation, and misconfigurations that could facilitate lateral movement. Review and harden inter-system authentication and access controls, particularly between web, cloud, and internal systems. Implement advanced monitoring solutions capable of detecting large-scale data exfiltration events and anomalous access patterns indicative of AI-driven attacks. Engage with national cybersecurity authorities such as CCN-CERT for incident response support and threat intelligence sharing.

The breach of Adif and Renfe underscores the urgent need for organizations to adapt their cybersecurity strategies to address the growing threat of AI-driven attacks. Traditional perimeter defenses and manual monitoring are increasingly insufficient in the face of adversaries leveraging automation and machine learning to accelerate the attack lifecycle. Proactive measures, including continuous vulnerability management, advanced behavioral analytics, and robust incident response capabilities, are essential to mitigate the risks posed by this new generation of cyber threats.

El Mundo, “Un grupo criminal hackea las webs de Adif y Renfe con una IA y roba 500 Gigabytes de datos”, 25 Sep 2026:

MITRE ATT&CK Framework:

CCN-CERT (Centro Criptológico Nacional):

Rescana is here for you

At Rescana, we are committed to helping organizations stay ahead of emerging cyber threats. Our Third-Party Risk Management (TPRM) platform empowers you to continuously monitor, assess, and mitigate risks across your digital supply chain, ensuring resilience against even the most sophisticated attacks. If you have any questions or require further information, our team is always available at [email protected].

us / Book a demo

Talk to Rescana this advisory, or book a demo of the platform.