Artificial intelligence is reshaping the supply and demand dynamics of the cybersecurity industry.
As a new generation of AI models rapidly identifies software vulnerabilities, the remediation workload facing corporate security teams has surged dramatically, forcing banks, energy companies and technology firms to expand cybersecurity budgets and hire more specialists.
At the same time, corporate demand for AI security tools is growing quickly, accelerating revenue growth at cybersecurity vendors such as Palo Alto Networks.
The surge in AI-driven vulnerability discovery leaves corporate security teams overwhelmed
The main challenge now facing the cybersecurity industry is that AI discovers software flaws far faster than companies can fix them.
Large language models can not only automatically scan vast amounts of software code, but also surface issues that once took security researchers considerable time to identify, rapidly increasing the number of potential threats companies must address.
Lee Klarich, chief product and technology officer at cybersecurity firm Palo Alto Networks, said that in May, when Anthropic gave about 12 technology and cybersecurity companies access to the Mythos model, roughly 1,500 open-source software vulnerabilities were reported globally that month.
By June, those same 12 companies had used Mythos to uncover more than 37,000 new vulnerabilities, though not all of them have been confirmed or publicly disclosed.
Data from research organization Epoch AI also shows that the number of high-severity software vulnerabilities reported in August and September alone exceeded the total for all of 2025.
These flaws could threaten sensitive corporate data and critical systems, greatly adding to the workload of corporate security teams.
A string of recent cybersecurity incidents has further heightened corporate anxiety.
Enterprise software startup Atomicwork, following multiple software security incidents, asked about 30 engineers to set aside other work and prioritize checking whether their own systems were affected.
Although the company ultimately confirmed it was not directly impacted, CEO Vijay Rayapati said the experience reflected how broadly unprepared companies are to handle such threats.
Banks and energy firms expand security budgets, and demand for AI security tools grows rapidly
Faced with mounting software vulnerabilities and the risk of cyberattacks, large enterprises are accelerating their cybersecurity investment.
Bank of America recently disclosed plans to hire 1,000 contract workers to use AI tools such as Anthropic's Mythos to fix software security flaws.
Goldman Sachs is also making extensive use of Mythos and OpenAI's Astra model to inspect internal software.
Goldman Sachs management has discussed that future spending on AI cybersecurity tools could match or even exceed spending on AI coding tools, which currently account for the bulk of the bank's AI expenditure.
The energy industry is also raising security budgets.
A major U.S. energy supplier plans to increase its cybersecurity budget of about $30 million this year by more than 10%, with the additional funds mainly used to buy AI security tools and upgrade existing security software such as Proofpoint and Cisco Splunk.
Worried that hackers will use AI to sustain attacks, the company has even drawn on part of next year's cybersecurity budget in advance.
Global IT services company Diversified plans to increase its cybersecurity software budget by about 20% over the next year and hire at least three full-time cybersecurity staff.
Previously, the company mobilized most of its roughly 100 IT employees to check and fix software vulnerabilities, which severely disrupted day-to-day operations.
Atomicwork also plans to roughly triple its number of cybersecurity engineers, expanding to 27 people over the next year.
Cybersecurity vendors gain growth opportunities as Palo Alto Networks' revenue growth doubles
Rising corporate security spending is directly driving growth in the cybersecurity industry.
As the world's largest independent cybersecurity vendor, Palo Alto Networks posted growth of more than 30% in its two most recently reported quarters, double its previous level.
Company executives believe that as AI continues to improve vulnerability discovery, enterprises must turn vulnerability detection, system monitoring and security remediation into ongoing daily work rather than handling them in a concentrated way only after an incident occurs.
The trend also creates new business opportunities for AI model developers.
OpenAI and Anthropic both view cybersecurity as an important source of revenue growth in the coming year.
Traditional security vendors such as Palo Alto Networks are integrating these companies' AI models into their own products, while startups that use AI to automate security work are beginning to attract interest from potential acquirers.
However, the cost of AI security tools is starting to become a concern for enterprises.
Scott Roberts, chief information security officer at enterprise software company UiPath, said that after rapid increases in spending on AI tools such as Mythos, the company recently shifted to lower-cost open-source models to reduce cybersecurity operating expenses.
Comments