Artificial intelligence is no longer just transforming how companies operate; it’s also reshaping the cybersecurity landscape. According to IBM’s report Cost of a Data Breach 2026, one in four recorded cyberattacks over the past year involved AI-based techniques, a 56% increase from the previous year. Furthermore, these breaches had an average cost of $6 million, about one million dollars more than the global average cost of a data breach, which stands at $4.99 million.
IBM Report Highlights in 30 Seconds
- One in four analyzed attacks utilized artificial intelligence, a 56% increase from the previous year.
- Breaches involving AI cost an average of $6 million, compared to $4.99 million for all incidents.
- More than 20% of organizations experienced targeted attacks on AI models or applications.
- Critical infrastructures accounted for 62% of AI-driven attacks.
- IBM notes that using AI and automation in security operations can reduce the cost of a breach by nearly $2 million.
The study, conducted by the Ponemon Institute and sponsored by IBM, analyzed 602 security breaches reported between March 2025 and February 2026 across organizations worldwide. The results reflect a significant shift in the cybercrime economy: while AI accelerates the time and costs needed to launch attacks, companies are investing increasingly more resources to detect, contain, and recover from them.
IBM suggests that this imbalance is changing the dynamics between attackers and defenders. Tools capable of generating phishing campaigns, creating fake identities with deepfakes, or developing malicious code with AI help automate tasks that previously required more time and specialized expertise.
AI as a Target
The report highlights another important trend: AI itself has become a target for cybercriminals.
More than 20% of participating organizations reported experiencing a breach directly affecting AI models or applications. However, IBM emphasizes that the issue usually lies not within the models themselves but in the surrounding systems.
The most common causes were:
- Compromised APIs (27%).
- Vulnerable applications or plugins (27%).
- Misconfigurations in cloud environments hosting AI workloads (27%).
These numbers indicate that many organizations are adopting AI services faster than they are updating their security controls.
Critical Infrastructure and Ransomware Among the Major Targets
Critical infrastructures accounted for 62% of the AI-driven attacks identified in the study.
Within these sectors, IBM highlights especially:
- Financial services, with an average incident cost of $6.3 million.
- Energy companies, with an average of $5.2 million.
The report also notes an increase in ransomware attacks. 39% of affected organizations reported experiencing such incidents, up from 34% the previous year.
Beyond encrypting systems, criminal groups increasingly use reputational pressure as a form of extortion. According to IBM, 41% of attacks aimed to damage the organization’s public image, ahead of data theft of employees (35%) or intellectual property (31%).
Automation Mitigates Economic Impact
Although AI facilitates attackers’ efforts, IBM points out that it can also be an effective defensive tool.
Organizations heavily utilizing AI and automation in their security operations reduced breach costs by nearly $2 million on average compared to those that haven’t adopted these technologies. Nevertheless, approximately one in four organizations still have not integrated such tools into their security operations centers.
The report also highlights other areas for improvement.
Only 37% of organizations encrypt sensitive data both at rest and in transit, and only 34% have full visibility over their cryptographic assets. IBM considers these shortcomings particularly critical as companies prepare for risks associated with quantum computing.
Businesses Accelerate Cybersecurity Investments
Complementing the main report, the Ponemon Institute conducted an additional survey among some of the participating organizations.
Eighty-five percent (85%) said they plan to increase their security investments after learning about the capabilities of advanced AI models, a higher figure than the 64% who initially believed they would only increase budgets after experiencing a security breach.
For IBM, this shift indicates that many companies are proactively reinforcing their defenses, including identity protection, threat detection automation, and vulnerability patching before exploits occur.
Frequently Asked Questions
How much does an AI-powered security breach cost on average?
According to IBM, the average cost is $6 million, roughly one million dollars more than the global average of all breaches analyzed.
What percentage of attacks involve artificial intelligence?
The report states that one in four malicious attacks employed AI-based techniques during the analyzed period.
Which sectors are most affected by AI-related attacks?
Critical infrastructure sectors, especially financial and energy, accounted for 62% of these incidents.
Can AI also help defend against cyber threats?
Yes. IBM asserts that organizations using AI and automation in their security operations reduce the average breach cost by nearly $2 million.
