What Does IBM’s 2025 Cost of a Data Breach Report Reveal About AI Risks, Shadow AI, and Security Solutions?

TL;DR

IBM’s 2025 report says extensive AI use in security can reduce breach identification and containment time by 80 days and costs by about $1.9 million. Based on 600 breached organizations, it puts the worldwide average breach cost at $4.44 million and highlights shadow AI, phishing, insider threats, and third-party risk. Read on for the report’s key figures and practical security priorities.

Transcript

How many times have you heard this? We really can't afford to spend that much on security. Maybe instead, we should be asking ourselves if we can afford not to. After all, it's not just about dollars. It's downtime. Reputation. Lost trust. And the fact is that many of these breaches are preventable. But let's face the decision regarding security in... Read More

Key Insights

• The worldwide average cost of a data breach is $4.44 million, a 9% decline from the prior figure discussed in the report. The calculation excludes exceptionally large mega breaches because those incidents would distort the average and make it less representative of typical organizational experiences.
• The combined time required to identify and contain a breach is 241 days, compared with about 257 days four or five years earlier. Most of that period is spent identifying the intrusion, leaving attackers inside affected environments for the better part of a year.
• The average United States data breach costs $10.22 million, following a 9% increase. That figure is almost twice the worldwide average, with higher regulatory fees and growing detection expenses identified as factors contributing to the increase.
• AI-related breaches affected 13% of surveyed organizations, and 60% of that group experienced data compromise while 31% experienced operational disruption. These findings show that AI implementations can introduce attack vectors with consequences for both sensitive information and business operations.
• Shadow AI is unauthorized artificial intelligence operating within an organization without approval or prior awareness. Twenty percent of organizations discovered these implementations in their environments, indicating that unmanaged AI can appear and remain unnoticed until it creates a problem.
• Insider threats produced the costliest attacks because insiders already understand the environment and can move directly toward valuable systems or information. Third-party risk was nearly tied in cost, including situations involving external access and third-party software.
• Phishing caused 16% of breaches and remained the most frequent attack vector discussed. AI can automate persuasive phishing content quickly, with a chatbot producing in five minutes a result that performed nearly as well as work created by a skilled cybersecurity professional over 16 hours.
• Extensive AI use in security reduced mean time to identify and contain breaches by 80 days and lowered associated costs by about $1.9 million. However, 63% of organizations had no AI governance policy or were still developing one, leaving many without defined objectives.

Questions & Answers

Q: How can organizations use AI security to reduce data breach costs?

Organizations using AI extensively in security reduced the combined time to identify and contain breaches by 80 days. Their breach costs were also about $1.9 million lower, showing that faster identification and containment can limit financial damage.

Q: What is the average cost of a data breach in IBM’s 2025 report?

The worldwide average cost is $4.44 million, down 9%. The report excludes exceptionally large mega breaches because they would distort the average.

Q: How long does it take to identify and contain a data breach?

The combined mean time to identify and contain a breach is 241 days, compared with about 257 days four or five years earlier. Most of that time is spent identifying the intrusion, so attackers can remain in an environment for much of a year.

Q: Why are data breaches more expensive in the United States?

The average breach cost in the United States rose 9% to $10.22 million, almost twice the worldwide average. Increased regulatory fees and detection expenses, including tooling and related work, contributed to the higher cost.

Q: What is shadow AI, and why is it a security risk?

Shadow AI refers to unauthorized AI implementations operating within an organization without approval or awareness. Twenty percent of organizations discovered shadow AI, creating visibility and governance problems because security teams cannot properly oversee systems they do not know exist.

Q: How often were AI-related breaches reported, and what damage did they cause?

AI-related breaches affected 13% of the surveyed organizations. Within that group, 60% experienced data compromise and 31% experienced operational disruption, demonstrating consequences for both information and business operations.

Q: Which data breach vectors were the most costly and frequent?

Insider threats caused the costliest incidents because insiders already understood the environment and could move directly toward valuable systems or information. Third-party risk was nearly tied in cost, while phishing was the most frequent initial vector at 16% of breaches.

Q: What security priorities does the 2025 report highlight for managing AI-related risk?

Organizations should combine AI-assisted security with stronger identity and access management, secrets management, employee defenses against phishing and deepfakes, and third-party risk controls. Formal AI governance is also important because 63% of organizations had no AI governance policy or were still developing one.

Summary & Key Takeaways

• IBM’s 2025 report draws on 600 organizations that experienced breaches and interviews with approximately 3,500 leaders who had direct knowledge of those incidents. The worldwide average breach cost fell 9% to $4.44 million, while the combined time needed to identify and contain a breach improved from about 257 to 241 days.

• The United States moved in the opposite direction, with its average breach cost rising 9% to $10.22 million. Increased regulatory fees and detection expenses contributed to the higher total. Insider threats and third-party risks produced the costliest incidents, while phishing remained the most frequent initial vector, accounting for 16% of breaches.

• AI creates risks and defensive opportunities. AI-related breaches affected 13% of organizations, while 20% discovered unauthorized shadow AI. Attackers used AI for convincing phishing and deepfakes, but organizations using AI extensively for security reduced identification and containment by 80 days and lowered breach costs by approximately $1.9 million.