AI In Cybersecurity Statistics By Market And Risks (2026)

Introduction

AI In Cybersecurity Statistics: Artificial intelligence has become both a powerful tool and a serious risk in the world of cybersecurity. While it helps defenders build stronger security systems, it also gives cybercriminals equally advanced tools to launch attacks. AI adoption has already become widespread, with more than two-thirds of IT and security professionals having tested AI tools. These tools are mainly used to monitor network traffic, create tests to check defenses, and predict potential security breaches before they happen.

Generative AI has made this situation even more complex. Security teams now use it to detect unusual behaviour and malicious activity faster and at a lower cost. However, cybercriminals are using the same technology to improve phishing attacks, develop new malware, and create more convincing deepfakes. In this article, we look at key statistics, real-world uses, and developing risks in cybersecurity.

Handpicked Feature by the Editors

1. The global AI cybersecurity market is expected to grow from USD 33 billion in 2025 to USD 163 billion by 2033, at a 22.3% CAGR.
2. Around 97% of organizations use AI-enabled security solutions.
3. Extensive AI use lowers the average breach cost to USD 3.62 million, compared with USD 5.52 million without AI.
4. Security AI and automation reduce breach-related costs by 34%.
5. Data leaks were the leading AI-related cybersecurity concern for 34% of surveyed leaders in 2026.
6. AI-powered spear-phishing achieved a 54% click-through rate while reducing costs by more than 95% versus manual methods.
7. Deepfake vishing attacks increased by 1,633% in Q1 2025 compared with Q4 2024.
8. Only 0.1% of people can accurately identify high-quality AI-generated deepfakes.
9. The AI cybersecurity sector completed 144 deals in 2025, while the solutions market reached USD 30.9 billion.

Global AI In Cybersecurity Market Statistics

• By 2025, the market reached USD 33 billion, followed by USD 40 billion in 2026 and USD 49 billion in 2027.
• The market continued rising to USD 60 billion in 2028 and USD 73 billion in 2029.
• In 2030, the market crossed USD 89 billion, growing further to USD 109 billion in 2031.
• By 2032, the market reached USD 134 billion, and it is expected to hit USD 163 billion by 2033.
• This growth reflects a strong yearly rate of 22.3%, known as CAGR, with Network Security remaining the leading product type throughout this period.

(Source: market.us)

AI Data Leaks Raise Cybersecurity Risks

• According to the Global Cybersecurity Outlook 2026 report by the World Economic Forum, data leaks caused by generative AI have become the top cybersecurity concern related to AI.
• More than 800 cybersecurity leaders were surveyed between August and October 2025. Among them, 34% identified data leaks as a major concern for 2026, ahead of hacker capability advancements at 29%.
• This marks a big change from previous years. In 2024 and 2025, more than 40% of experts were concerned about attacker innovations, while only around 20% focused on data exposure risks.
• The report’s authors describe this shift as a turning point in AI-related risks for the coming year, as generative AI tools become more common in daily tasks, increasing risks of data leaks and cyber espionage.
• Other concerns include the technical security of AI systems, rising from 5% in 2025 to 13% in 2026, and increased complexity of security governance, growing from 12% to 13% during the same period.

(Source: statista.com)

Impact of GenAI on Cybersecurity Attacks

• According to the 2025 Gartner Cybersecurity Innovations in AI Risk Management and Use Survey, organizations reported varying levels of impact from GenAI-related cyberattacks over the past 12 months.
• For attacks on enterprise GenAI application infrastructure, 71% of respondents reported no incidents, while 26% experienced at least one minor incident, and 3% faced a major incident.
• Regarding attacks on AI applications using the application prompt, 68% reported no incidents, 28% had minor incidents, and 4% experienced major incidents.
• For deepfake video used against automated face verification, 69% reported no incidents, while 27% faced minor incidents, and 3% experienced major incidents.
• In the case of deepfake audio used against automated voice biometrics, 67% reported no incidents, 30% had minor incidents, and 2% faced major incidents.
• For social engineering attacks using deepfakes during video calls with employees, 63% reported no incidents, 31% had minor incidents, and 5% experienced major incidents.
• Social engineering using deepfake audio during employee calls showed the highest impact, with 56% reporting no incidents, 38% facing minor incidents, and 6% experiencing major incidents.
• Major incidents in this survey were defined as those resulting in loss of intellectual property, significant financial loss, or business interruption.

(Source: gartner.com)

Major Cybersecurity Skills Gaps

• Cloud security is the largest skills gap, reported by 35% of cybersecurity professionals.
• AI and machine learning rank second, with 32% reporting a shortage of these skills.
• Zero-trust implementation skills are lacking at 29% of organizations.
• Penetration-testing skills are in short supply at 27% of organizations.
• App security skills are missing at 26% of organizations.
• Digital forensics and incident-response skills also have a 26% gap.
• Risk assessment, analysis, and management have a 24% skills gap.
• Security engineering skills are lacking in 23% of organizations.
• Threat intelligence analysis also has a 23% skills gap.
• Malware research and analysis rank lowest in the chart but still have a 22% gap.

(Reference: statista.com)

Benefits of Generative AI in Cybersecurity

• Cloud security leads the list, with 55% of respondents expecting significant benefits from generative AI.
• Security operations and management rank second at 52%.
• Endpoint security also receives support from 52% of respondents.
• Email security and security awareness are expected to benefit, according to 50% of respondents.
• Identity and access management receives 47% support.
• Data protection is selected by 46% of respondents.
• Application security is expected to benefit according to 43%.
• Network security receives support from 42% of respondents.
• Web security is selected by 31%.
• Governance, risk, and compliance receive 28% support.
• Internet of Things and operational technology security are selected by 23% of respondents.
• Security consulting, advisory services, and assessments receive 14% support.
• Managed security service provider outsourcing ranks last, with 9% expecting significant benefits.

(Source: mckinsey.com)

Benefits of AI in Cybersecurity

• Security AI and automation can deliver average cost savings of USD 1.9 million.
• Organizations using security AI can reduce breach-related costs by 34%.
• The average cost of a data breach with extensive AI use is USD 3.62 million.
• Without AI, the average cost of a data breach rises to USD 5.52 million.
• AI and automation can help organizations detect a data breach within 51 days.
• Companies can achieve annual security cost savings of approximately USD 2.22 million by adopting AI.
• AI can reduce the time required to respond to security incidents by 80 days.
• Around 97% of organizations use AI-enabled security solutions.
• Approximately 77% of security teams are adopting AI at a steady pace.
• Around 75% of cybersecurity professionals use AI tools in their work.

(Source: stationx.net)

AI Security Spending Stats

• 41% of respondents selected observability, including model monitoring and logs, as a top-three AI security priority.
• 35% chose AI governance, such as maintaining a catalogue of AI services.
• 32% prioritised protection against sensitive-data leaks and personally identifiable information exposure.
• 28% selected production vulnerability monitoring for AI models.
• 26% prioritised pre-production code scanning for AI models, and 23% chose measures to reduce data-poisoning risks.
• Another 23% prioritised continuous “red teaming” to test AI models for weaknesses.
• 22% selected protection against denial-of-service attacks targeting AI models.
• 19% prioritised monitoring for model drift and declining model quality.
• 17% selected protection against input manipulation, including prompt-injection attacks.
• For additional third-party security spending, 2.5% of respondents expected no extra costs.
• 45.6% expected additional spending of less than 5%, making this the largest group.
• 17.7% expected costs to rise by 5% to less than 10%, another 17.7% expected additional spending of 10% to less than 15%.
• 16.5% expected spending to increase by at least 15%.

(Source: mckinsey.com)

AI Deepfake and Vishing Attack Statistics

• AI-powered deepfakes and vishing, or voice phishing, are among the fastest-growing cyberattack methods.
• Deepfakes were used as an attack method in 35% of data breaches.
• Vishing attacks increased by 442% from the first half to the second half of 2024.
• Deepfake-related fraud rose by 1,300% during 2024.
• Deepfake vishing attacks surged by 1,633% in the first quarter of 2025 compared with the fourth quarter of 2024.
• The largest confirmed deepfake scam involved a fake chief financial officer video call and resulted in a transfer of USD 25.6 million.
• Fraud losses linked to generative AI are projected to reach USD 40 billion by 2027.
• Only 0.1% of people can accurately identify high-quality AI-generated deepfakes.
• Around 21% of managers said their organizations were least prepared to handle deepfake attacks.

AI Phishing Attack Statistics

• AI-powered spear-phishing emails achieve a 54% click-through rate, similar to the results produced by experienced human red-team specialists.
• AI can reduce the cost of spear-phishing campaigns by more than 95% compared with manual methods.
• Around 82.6% of phishing emails now use AI to improve their content, targeting, or delivery.
• In the second quarter of 2024, approximately 40% of business email compromise messages were generated using AI.
• AI-generated phishing was used as an attack method in 37% of data breaches covered by the 2025 report.
• AI phishing performance improved by 55% between 2023 and 2025.
• Around 91% of security professionals reported facing AI-powered email attacks during the previous 6 months.
• More than 80% of social-engineering attacks now involve AI-supported methods.

AI Cybersecurity Companies and Vendors

• In 2025, the AI cybersecurity sector recorded 144 deals, making it the most active cybersecurity investment category.
• Around 77% of security teams were adopting AI at pace, while 75% of security professionals were using AI tools.
• AI-enhanced SIEM and XDR platforms received 31% of cybersecurity budgets.
• Endpoint detection and response solutions accounted for another 19% of cybersecurity budgets.
• The AI cybersecurity solutions market was valued at USD 30.9 billion in 2025.
• CrowdStrike offers Falcon and Charlotte AI for natural-language threat hunting across EDR, XDR, and SIEM.
• Palo Alto Networks provides Cortex XSIAM for an AI-driven security operations center and platform consolidation.
• Microsoft offers Security Copilot for GPT-4-powered investigations within Azure and Microsoft 365.

Recent AI In Cybersecurity Developments

• On September 17, 2026, AI-powered security and compliance startup Comp AI raised USD 34 million in Series A funding, bringing its total funding to USD 37.5 million.
• On September 1, 2026, AI security startup AIR launched from stealth with USD 50 million raised through two seed rounds to protect the software supply chain used by AI agents.
• On August 13, 2026, IBM and OpenAI expanded their partnership to combine OpenAI models with IBM Autonomous Security and train tens of thousands of consultants in secure enterprise AI deployment.
• On August 5, 2026, Sophos partnered with OpenAI to integrate frontier models into Sophos Fusion and provide AI-security services to managed service providers.
• On August 3, 2026, Horizon3.ai raised USD 250 million in Series E funding at a USD 2 billion valuation for its autonomous penetration-testing platform.
• On July 22, 2026, endpoint-security startup Glow emerged from stealth with USD 180 million in Series A funding at a USD 1.2 billion valuation.
• On June 18, 2026, Accenture announced cybersecurity acquisitions worth USD 4.18 billion, including stakes in Dragos and full ownership of runZero and NetRise, to expand its USD 10 billion security business.
• On March 10, 2026, Mandiant founder Kevin Mandia launched AI-native cybersecurity startup Armadin with USD 189.9 million in combined seed and Series A funding.

Summary

AI-driven defence has proven to be financially valuable, cutting average data breach costs by over one-third and saving companies millions of dollars each year. However, the same technology is also being used by attackers to become more effective. Many companies still face a shortage of skilled workers in AI and machine learning.

Most people also struggle to identify deepfakes, and spending on AI security tools varies greatly between organizations. These gaps are slowing down how quickly companies can strengthen their defences. The focus is shifting from simply using AI as a defence tool to also monitoring and controlling AI systems themselves. This includes improving visibility into how AI operates and preventing sensitive data from leaking through these systems.

FAQ