More than 20% of organizations reported a breach targeting AI models or applications

ARMONK, N.Y., July 29, 2026 /PRNewswire/ -- One in four malicious breaches were AI-enabled – a 56% increase over last year – and these breaches cost an average of $6 million, roughly $1 million more than the global breach average of $4.99 million, according to IBM's 2026 Cost of a Data Breach Report.

These attacks, compromised of mostly deepfake impersonation and AI-enabled malware, are reshaping breach economics. Attacks are getting faster and cheaper to launch, while breaches keep getting more expensive to find and fix. Companies that reported using AI and automation in security operations cut breach costs by an average of almost $2 million dollars, yet one in four organizations have still not adopted these tools in their security operations.

This growing imbalance—where attacks can be launched for thousands while breaches cost millions—is fundamentally changing the economics of cyber risk.

Frontier AI Threats Driving Earlier Action

Organizations are starting to act on future risk, rather than waiting for an incident. In separate follow-on research conducted by Ponemon Institute, 85% of organizations said they plan to increase security spending after becoming aware of advanced frontier AI cyber capabilities – compared to just 64% that reported in the initial research that they plan to increase security spend after experiencing a breach.

But a gap remains where attackers are moving fastest. While more than 50% reported using agents for threat detection and containment, only 18% apply agents to vulnerability management, leaving known exposures to linger even as AI shortens exploit windows. Three quarters of organizations say frontier AI threats are prompting them to rethink how agents are deployed across their security operations.

"What's changing is the economics of cyberattacks. AI is making attacks faster and cheaper, while breaches keep getting more expensive. When organizations have an extended gap between discovery and remediation, that imbalance shows up directly in breach costs," said Suja Viswesan, VP, IBM Security Software. "The priority now is to eliminate that lag—building remediation into development workflows, securing identity at runtime, and fixing risks at the speed attackers are already moving."

Critical Infrastructure Face Higher AI-Driven Risk

Most AI-driven attacks reported in the study targeted critical infrastructure sectors (62%), with financial services and energy organizations experiencing the highest concentration, raising the risk of broader systemic disruption. Financial services breaches were reported to cost  on average $6.3 million, while energy breaches cost on average $5.2 million. The concentration of attacks across these sectors increases the potential for cascading impacts across economies, supply chains, and essential services.

Other Key Findings:

• AI's Weakest Link. More than 20% of organizations reported a breach targeting AI models or applications. The most common causes were weaknesses in surrounding systems: compromised APIs, applications, or plug-ins (27%) and cloud misconfigurations affecting AI workloads (27%).
• Encryption Gaps Persist as Quantum Risk Looms. Core weaknesses in encryption and cryptographic management continue to expose organizations, even as quantum-safe investments grow. Only 37% of breached organizations stated that they encrypt sensitive data both at rest and in transit, and just 34% have visibility into cryptographic assets.
• Ransomware Actors Weaponize Reputation. Reported ransomware incidents rose compared to the year prior (39% vs. 34%), with attackers increasingly using AI to automate and scale. While operational disruption still plays a role, attackers are shifting toward higher-impact pressure—most commonly exploiting brand reputation (41%), followed by employee data (35%) and intellectual property (31%).

The 2026 report, conducted by Ponemon Institute and sponsored and analyzed by IBM, is based on breaches experienced by 602 organizations globally between March 2025 and February 2026. The follow-on study was conducted in May 2026, where 456 organizations of the 602 from the CODB research responded. Of these organizations, 78% or 356 of organizations were aware of recent reports about highly advanced frontier models such as Mythos. 

Additional Resources

• Download the full report.
• Register for the webinar.
• Schedule a briefing with IBM experts.
• Understand your AI Cyber Resilience.

About IBM

IBM (NYSE: IBM) is a leading provider of global hybrid cloud and AI, and consulting expertise. We help clients in more than 175 countries capitalize on insights from their data, streamline business processes, reduce costs and gain the competitive edge in their industries. Thousands of governments and corporate entities in critical infrastructure areas such as financial services, telecommunications and healthcare rely on IBM's hybrid cloud platform and Red Hat OpenShift to affect their digital transformations quickly, efficiently and securely. IBM's breakthrough innovations in AI, quantum computing, industry-specific cloud solutions and consulting deliver open and flexible options to our clients. All of this is backed by IBM's long-standing commitment to trust, transparency, responsibility, inclusivity and service. Visit www.ibm.com for more information.

Media Contact

IBM

Michele Brancati

mbrancati@ibm.com

View original content to download multimedia:https://www.prnewswire.co.uk/news-releases/ibm-study-one-in-four-malicious-breaches-are-ai-enabled-costing-companies-6-million-on-average-302837055.html

Der US-Datenkonzern Palantir hat in der Schweiz eine deutliche juristische Niederlage erlitten. Das Handelsgericht Zürich wies 22 von 23 beanstandeten Passagen in zwei kritischen Artikeln des Online-Magazins «Republik» zurück. Lediglich eine einzelne Formulierung muss die Redaktion korrigieren. Palantir und seine Schweizer Tochtergesellschaft wollten gerichtlich durchsetzen, dass umfangreiche Gegendarstellungen publiziert werden; dieses Anliegen ist nun weitgehend gescheitert. Nach dem Urteil muss der Konzern zudem den überwiegenden Teil der Verfahrenskosten tragen.

Ausgangspunkt des Rechtsstreits waren Recherchen, die im Dezember 2025 nach rund einjähriger Arbeit erschienen. Das Projekt von «Republik» und dem Recherchekollektiv WAV beleuchtete, wie Palantir über mehrere Jahre versucht hatte, bei Bundesstellen und dem Schweizer Militär Aufträge zu erhalten – ohne Erfolg. Die Journalisten werteten dazu Dutzende Gesuche nach dem Öffentlichkeitsgesetz sowie interne Unterlagen aus und kamen zum Befund, dass das Unternehmen trotz einer mehrjährigen Präsenz in der Schweiz keinen einzigen Regierungsauftrag gewinnen konnte. Diese Erzählung eines «Scheiterns» auf dem Schweizer Markt war nach Darstellung der beteiligten Reporter Auslöser für Palantirs juristische Offensive.

Die Redaktion hatte dem Unternehmen nach eigenen Angaben vor Veröffentlichung ausführliche Fragen geschickt und Führungskräfte interviewt. Palantir verlangte darüber hinaus eine detaillierte, punktweise Replik im Magazin, die über den Umfang der ursprünglichen Recherche hinausgegangen wäre. Als «Republik» dies ablehnte, reichte der Konzern Klage ein. Das Zürcher Handelsgericht stellte nun klar, dass nur eine von 23 umstrittenen Passagen eine Gegendarstellung erfordert. Damit bestätigt das Urteil, dass der Großteil der Berichterstattung über Palantirs vergebliche Akquisitionsversuche bei Schweizer Behörden auf einer zulässigen Auswertung amtlicher Dokumente beruht.

Palantir ist international vor allem durch den Einsatz seiner Software in militärischen und nachrichtendienstlichen Kontexten bekannt. Seit 2022 stellt das Unternehmen Plattformen zur Verfügung, die ukrainische Behörden und Streitkräfte bei der Auswertung von Lagebildern, der Zielidentifikation und der Logistik unterstützen. Die Systeme werden außerdem von US-amerikanischen und verbündeten Diensten im Nahen Osten genutzt, um iranische Aktivitäten etwa in Raketenprogrammen, beim Einsatz von Drohnen oder in Cyberoperationen zu analysieren. Direkte Geschäftsbeziehungen mit dem Iran bestehen nach Unternehmensangaben aufgrund von US-Sanktionen nicht. Das Zürcher Urteil könnte über die Schweiz hinaus Signalwirkung haben, nachdem die Recherche bereits in anderen europäischen Staaten Debatten über Nutzen und Notwendigkeit der Palantir-Technologie im staatlichen Bereich angestoßen hatte.