Hey there! Looking for smarter insights delivered straight to your inbox? Sign up for our weekly newsletters tailored for enterprise AI, data, and security leaders. Don’t miss out – subscribe now!
Did you know that Shadow AI is the $670,000 problem that many organizations are unaware of?
IBM’s latest report on the Cost of a Data Breach for 2025, done in collaboration with the Ponemon Institute, sheds light on the fact that breaches related to employees’ unauthorized use of AI tools are costing organizations an average of $4.63 million. This is a significant 16% higher than the global average breach cost of $4.44 million.
The research, based on interviews with 3,470 individuals from 600 breached organizations, highlights the rapid adoption of AI without adequate security oversight. While only 13% of organizations reported AI-related security incidents, a staggering 97% of those breached lacked proper AI access controls. Additionally, 8% were unsure if they had been compromised through AI systems.
According to Suja Viswesan, Vice President of Security and Runtime Products at IBM, “The data indicates a gap between AI adoption and oversight already exists, with threat actors starting to exploit it. The report uncovers a lack of basic access controls for AI systems, leaving sensitive data exposed and models vulnerable to manipulation.”
Exciting News! The AI Impact Series Returns to San Francisco on August 5
The next phase of AI is here – are you ready? Join leaders from Block, GSK, and SAP for an exclusive look at how autonomous agents are reshaping enterprise workflows – from real-time decision-making to end-to-end automation.
Reserve your spot now – space is limited: https://bit.ly/3GuuPLF
Shadow AI and Supply Chains: A Favorite Target for Attackers
The report reveals that 60% of AI-related security incidents led to compromised data, with 31% causing disruptions to daily operations. Personal identifiable information (PII) of customers was compromised in 65% of shadow AI incidents, significantly higher than the global average of 53%. Governance remains a weakness, with 63% of breached organizations lacking AI governance policies or still in the process of developing them.
As Itamar Golan, CEO of Prompt Security, puts it, “Shadow AI is akin to doping in the Tour de France; seeking an edge without considering the long-term consequences.” Prompt Security has identified over 12,000 AI apps and detects 50 new ones daily.
Adversaries continue to outpace defenses against software and model supply chain attacks. The report identifies supply chains as the primary attack vector for AI security incidents, with 30% involving compromised apps, APIs, or plug-ins.
Weaponized AI on the Rise
Various forms of weaponized AI, including LLMs designed to enhance tradecraft, are rapidly proliferating. Sixteen percent of breaches now involve attackers leveraging AI, primarily for AI-generated phishing (37%) and deepfake attacks (35%). Purpose-built models like FraudGPT, GhostGPT, and DarkGPT are available for as little as $75 a month and are used for attack strategies such as phishing, exploit generation, and code obfuscation.
Cisco’s The State of AI Security Report indicates that fine-tuned LLMs are 22 times more likely to produce harmful outputs than base models. “Adversaries are not just automating attacks with AI; they’re using it to blend into normal network traffic, making detection more challenging,” says Etay Maor, Chief Security Strategist at Cato Networks.
Shlomo Kramer, CEO of Cato Networks, warns, “Companies have a limited window to avoid being caught off guard by fragmented architectures. Attackers are moving faster than integration teams.”
Governance: A Weakness Exploited by Adversaries
Despite 37% of organizations claiming to have AI governance policies, only 34% conduct regular audits for unsanctioned AI, and a mere 22% perform adversarial testing on their AI models. DevSecOps emerges as a key factor in reducing breach costs, saving organizations an average of $227,192.
The report emphasizes how neglecting governance impacts long-term security, with 63% of breached organizations lacking AI governance policies or still developing them. Chris Goettl, VP Product Management for Endpoint Security at Ivanti, stresses the importance of timely vulnerability management over traditional patch management.
The $1.9M AI Dividend: Smart Security Pays Off
Amidst the rise of weaponized AI, the report provides hope for combating evolving threats. Organizations extensively using AI and automation save $1.9 million per breach and resolve incidents 80 days faster. These AI-powered organizations spend $3.62 million on breaches, compared to $5.52 million for those without AI, showcasing a 52% cost differential.
Vineet Arora, CTO at WinWire, highlights how AI’s data analysis capabilities transform security economics, enabling extensive AI users to operate 18% below the global average breach cost. However, adoption of AI security remains a challenge, with only 32% extensively using AI security tools.
Security teams are keeping pace, with 77% matching or surpassing their company’s overall AI adoption. Post-breach investments see 45% opting for AI-driven solutions, focusing on threat detection, incident response planning, and data security tools.
Combining DevSecOps practices with AI solutions further reduces breach costs by over $2 million, positioning security as a competitive differentiator rather than just a cost center.
Understanding U.S. Cybersecurity Costs vs. Global Trends
The cybersecurity landscape in 2024 showcased a paradox: while global breach costs decreased to $4.44 million, U.S. organizations faced record-high costs of $10.22 million per incident. Healthcare organizations bore the brunt, with an average breach cost of $7.42 million and extended resolution timelines of 279 days.
Operational disruptions were equally severe, with 86% of breached organizations reporting significant business impact. Investment fatigue post-breach saw a decline from 63% to 49% year-over-year, indicating doubts regarding the ROI of reactive security investments. Resilience strategies need reevaluation, especially for healthcare providers facing prolonged recovery times.
IBM’s Call for Strong Governance
George Kurtz, CEO of CrowdStrike, notes how Gen AI has empowered cybercriminals, while Mike Riemer, Field CISO at Ivanti, predicts a shift in defenders leveraging AI for cybersecurity in 2025.
The report highlights immediate actions organizations can take:
- Implement AI governance now – Only 45% have approval processes for AI deployments.
- Gain visibility into shadow AI – Regular audits are crucial given that 20% experience breaches from unauthorized AI.
- Accelerate security AI adoption – The $1.9 million in savings justifies proactive deployment.
The report concludes, emphasizing collaboration between key stakeholders to ensure integrated security and governance practices. In a landscape where AI-powered attacks escalate, organizations must balance AI benefits with rigorous risk management for survival.
