16
Artificial intelligence is making cyberattacks faster, cheaper and easier to scale, while also giving organisations new tools to detect vulnerabilities and respond to threats, according to a
new Mastercard report.
The report found that the average cost of independently reproducing a working software exploit has fallen to about $2.77, as widely available AI models reduce the technical expertise and time required to exploit known vulnerabilities. AI systems have also improved their performance on cybersecurity tasks, with success rates on apprentice-level tasks rising from 9% to 50% over two years.
The shift is creating a more complex cybersecurity environment for businesses, as attackers can use AI to identify vulnerabilities and automate parts of the attack process.
At the same time, AI is being used to strengthen cyber defences. Mastercard said AI can now identify 86% of vulnerabilities created by AI, compared with 37% previously. The technology is also helping security teams detect scams and impersonation attempts more quickly.
The report highlights the growing importance of responding to cyber threats at machine speed, as the time available to identify, investigate and patch vulnerabilities continues to shrink.
In India, the technology, public and financial sectors together account for about 44% of cyberattacks, according to data cited in Mastercard's report. Business systems and customer personal and financial information also represent more than half of the assets targeted in attacks.
The rise of AI agents is adding another layer of risk. These systems can perform tasks and transactions on behalf of users, but questions remain around how their identities should be verified and who would be responsible when an AI agent behaves unexpectedly.
Mastercard said organisations will need to combine AI-powered security tools with stronger governance, monitoring and response mechanisms as cyber threats increasingly operate at machine speed.
The findings come as businesses globally increase their use of AI while facing a parallel rise in automated cyber threats, making the ability to detect and respond to attacks a key part of AI adoption.
The report found that the average cost of independently reproducing a working software exploit has fallen to about $2.77, as widely available AI models reduce the technical expertise and time required to exploit known vulnerabilities. AI systems have also improved their performance on cybersecurity tasks, with success rates on apprentice-level tasks rising from 9% to 50% over two years.
The shift is creating a more complex cybersecurity environment for businesses, as attackers can use AI to identify vulnerabilities and automate parts of the attack process.
At the same time, AI is being used to strengthen cyber defences. Mastercard said AI can now identify 86% of vulnerabilities created by AI, compared with 37% previously. The technology is also helping security teams detect scams and impersonation attempts more quickly.
The report highlights the growing importance of responding to cyber threats at machine speed, as the time available to identify, investigate and patch vulnerabilities continues to shrink.
In India, the technology, public and financial sectors together account for about 44% of cyberattacks, according to data cited in Mastercard's report. Business systems and customer personal and financial information also represent more than half of the assets targeted in attacks.
The rise of AI agents is adding another layer of risk. These systems can perform tasks and transactions on behalf of users, but questions remain around how their identities should be verified and who would be responsible when an AI agent behaves unexpectedly.
Mastercard said organisations will need to combine AI-powered security tools with stronger governance, monitoring and response mechanisms as cyber threats increasingly operate at machine speed.
The findings come as businesses globally increase their use of AI while facing a parallel rise in automated cyber threats, making the ability to detect and respond to attacks a key part of AI adoption.
