IBM Cost of a Data Breach Report 2025

Average Cost of a Data Breach: 2025 Statistics

The most comprehensive data breach cost dataset available. All figures from IBM's analysis of 604 organizations across 17 countries.

Global Average

$4.44M

Down 9% from 2024

US Average

$10.22M

Record high, +9% YoY

Healthcare

$7.42M

#1 for 15 consecutive years

Detection Time

241 days

Lowest in 9 years

Year-Over-Year Cost Trend (2019-2025)

After reaching an all-time high of $4.88M in 2024, global breach costs declined 9% to $4.44M in 2025 -- the first significant decrease in four years. Analysts attribute this decline to increased adoption of AI-powered security tools and faster detection capabilities. However, this global average masks regional divergence: US costs rose 9% to $10.22M, reflecting increasing regulatory complexity and litigation costs.

2019
$3.92M
2020
$3.86M
2021
$4.24M
2022
$4.35M
2023
$4.45M
2024
$4.88M
2025
$4.44M

Source: IBM Cost of a Data Breach Report 2019-2025 (Ponemon Institute)

Cost Per Record by Data Type

Not all data is valued equally in a breach. Intellectual property carries the highest per-record cost at $178, reflecting the long-term competitive damage and potential regulatory liability when proprietary information is exposed. Customer personally identifiable information (PII) costs $160 per record, driven by notification requirements, credit monitoring obligations, and class action exposure. Employee PII, while slightly less expensive at $156 per record, still represents significant liability due to employment law implications and internal trust damage. Even anonymized data carries a $130 per-record cost, as re-identification risks and regulatory scrutiny around anonymization techniques have increased.

Intellectual Property$178/record
Customer PII$160/record
Employee PII$156/record
Anonymized Data$130/record

Source: IBM Cost of a Data Breach Report 2025

Detection and Containment Timeline

The mean time to identify and contain a data breach dropped to 241 days in 2025, the lowest figure in nine years of IBM's research. This improvement correlates strongly with increased adoption of AI-powered security monitoring, which reduced the breach lifecycle by an average of 80 days. The cost implications of detection speed remain stark: organizations that identified and contained breaches within 200 days spent $3.87 million on average, while those exceeding 200 days faced costs of $5.01 million -- a 23% premium of $1.14 million for slower detection. This finding makes detection speed one of the most impactful variables in breach cost, after geographic region and industry sector.

Mean Detection Time

241 days

Lowest in 9 years

Under 200 Days

$3.87M

Fast detection saves money

Over 200 Days

$5.01M

23% more expensive

The 200-day threshold is not arbitrary. IBM's research consistently shows a cost inflection point around this mark. Breaches detected quickly tend to be contained before attackers achieve lateral movement, exfiltrate large data volumes, or establish persistent access. After 200 days, the probability of data exfiltration, regulatory notification triggers, and customer churn all increase significantly, compounding costs across every category. Credential-based attacks take the longest to detect (292 days on average), while internally-detected breaches through security tools are identified fastest.

Source: IBM Cost of a Data Breach Report 2025

Cost by Attack Vector

Ransomware remains the most expensive attack vector at $5.08 million per breach, reflecting the combined costs of ransom demands, extended downtime, forensic investigation, and recovery operations. Malicious insider attacks ($4.92M) are nearly as costly due to their difficulty to detect and the extensive access insiders possess. Business email compromise ($4.88M) continues its upward trend, driven by increasingly sophisticated social engineering that bypasses traditional security controls. Phishing ($4.76M) and credential theft ($4.67M) remain the most common attack vectors by volume. Cloud misconfiguration ($4.14M) carries the lowest average cost among tracked vectors, partly because these breaches tend to be detected faster through cloud security posture management tools.

Ransomware$5.08M
Malicious Insider$4.92M
Business Email Compromise$4.88M
Phishing$4.76M
Credential Theft$4.67M
Cloud Misconfiguration$4.14M

Source: IBM Cost of a Data Breach Report 2025, Verizon DBIR 2025

AI Impact on Data Breach Costs

The 2025 report reveals a growing divide between organizations that have embraced AI-powered security and those that have not. Companies with extensive AI and automation deployment saved $1.9 million per breach on average compared to those without -- the largest technology-driven cost difference IBM has ever measured. AI-powered security tools reduced the breach lifecycle by 80 days on average, translating directly into lower costs across all four IBM cost categories. The most significant AI benefits appear in detection and escalation, where automated threat detection, log analysis, and anomaly identification dramatically accelerate the initial response.

AI/Automation Savings

$1.9M

Per breach on average

Shadow AI Cost

+$0.67M

Unauthorized AI tool risk

Lifecycle Reduction

80 days

Faster detection & containment

The shadow AI risk is new for 2025. As employees adopt unauthorized AI tools (ChatGPT, Copilot, Gemini) for work tasks, they often input sensitive data into systems outside the organization's security perimeter. IBM found that breaches involving shadow AI cost an additional $670,000 on average, driven by expanded attack surface, data leakage through AI model inputs, and the difficulty of detecting unauthorized tool usage. This finding creates a new category of risk that most organizations have not yet addressed in their security policies.

The organizations benefiting most from AI security are those with mature implementations integrated into their security operations centres (SOCs). Simply purchasing AI-branded tools without proper integration yields minimal benefit. The key capabilities driving savings include automated alert triage (reducing false positive investigation time by 90%+), AI-assisted incident investigation (correlating indicators across data sources in seconds rather than hours), and predictive risk scoring that prioritizes vulnerabilities most likely to be exploited.

Source: IBM Cost of a Data Breach Report 2025

Understanding IBM's Methodology

The IBM Cost of a Data Breach Report is conducted annually by the Ponemon Institute using activity-based costing (ABC) methodology. The 2025 report analyzed 604 organizations that experienced real data breaches between March 2024 and February 2025, across 17 countries and 16 industries. This is not a survey of hypothetical costs -- it measures actual expenditures incurred by organizations that suffered breaches, making it the most empirically grounded study of its kind.

The study categorizes costs into four areas: detection and escalation (forensics, investigation, audit, crisis management), notification (contacting affected individuals and regulators), post-breach response (help desk, credit monitoring, legal, identity protection), and lost business (customer churn, revenue loss during downtime, reputation damage, diminished goodwill). Each cost is tracked over two years following the breach, recognizing that many expenses -- particularly litigation and customer churn -- extend well beyond the initial incident.

It is important to note several limitations. The sample skews toward larger organizations (the average breach involved millions of records), meaning small business costs may differ significantly. The study also relies on estimates from organizational representatives, which may not capture all hidden costs such as opportunity cost, executive distraction, or long-term competitive damage. Finally, mega-breaches (over 1 million records) are analyzed separately from the main sample, so the $4.44M average excludes the most extreme incidents.

Source: IBM Cost of a Data Breach Report 2025. All figures verified: April 2026.

Related Resources

Breach Cost Calculator

Calculate your specific breach exposure using IBM 2025 data.

Cost by Industry

Healthcare, financial, tech -- compare all 10 industry sectors.

Cost by Country

US ($10.22M) to Brazil ($1.36M) -- regional cost comparison.

Ransomware Deep Dive

$5.08M average. Payment economics and recovery costs.

Where the Money Goes

Lost business (38%), detection (29%) -- the full anatomy.

Prevention ROI

10 security controls ranked by return on investment.

Frequently Asked Questions

The global average cost of a data breach in 2025 is $4.44 million, according to IBM's Cost of a Data Breach Report 2025. This represents a 9% decrease from 2024's record high of $4.88 million. However, costs vary dramatically by region (US: $10.22M, Middle East: $7.29M, Brazil: $1.36M) and industry (Healthcare: $7.42M, Financial: $5.56M, Government: $2.83M). The report analyzed 604 real breaches across 17 countries, making it the most comprehensive annual study of breach costs available.
Per-record costs vary by data type. Intellectual property is the most expensive at $178 per record, reflecting long-term competitive damage and regulatory exposure. Customer PII costs $160 per record, driven by notification requirements, credit monitoring obligations, and litigation risk. Employee PII costs $156 per record, while anonymized data costs $130 per record. These figures represent averages across all industries -- healthcare records cost significantly more ($408/record) due to HIPAA requirements and the sensitivity of medical information.
The IBM Cost of a Data Breach Report is an annual study conducted by the Ponemon Institute (an IBM subsidiary) that measures the actual costs incurred by organizations experiencing data breaches. The 2025 report analyzed 604 organizations across 17 countries and 16 industries that suffered real breaches between March 2024 and February 2025. It uses activity-based costing methodology to track expenses across four categories: detection and escalation, notification, post-breach response, and lost business. First published in 2006, it is considered the gold standard for breach cost benchmarking.
Data breach costs have generally trended upward over the past decade, though with some fluctuation. In 2019, the global average was $3.92M. It dipped slightly in 2020 ($3.86M) during COVID, then climbed steadily: $4.24M (2021), $4.35M (2022), $4.45M (2023), reaching a record $4.88M in 2024. The 2025 figure of $4.44M represents the first significant decline in four years, attributed to improved AI-powered detection and faster response capabilities. Despite this global decrease, US costs rose 9% to $10.22M, reflecting increasing regulatory complexity.
According to IBM's research, only 53% of data breach costs are incurred in the first year. An additional 24% of costs emerge in year two, primarily from ongoing litigation, continued customer churn, and regulatory proceedings. The remaining 23% of costs occur in year three and beyond, driven by class action settlements, long-term brand damage, and increased compliance requirements. Equifax provides a clear example: breached in 2017, the company's total costs exceeded $1.4 billion through 2025, with major settlements and remediation expenses continuing eight years after the initial incident.
Yes, significantly. IBM's 2025 report found that organizations with extensive AI and security automation deployment saved $1.9 million per breach on average compared to those without these capabilities. AI tools reduced the breach lifecycle by 80 days through faster threat detection, automated alert triage, and accelerated investigation. However, AI also introduces new risks: shadow AI (unauthorized use of AI tools by employees) added $670,000 to breach costs on average. The key to realizing AI's benefits is mature integration into security operations, not simply purchasing AI-branded tools.