AI agents to play key role in ANZ IT security, report finds

The latest Salesforce State of IT report indicates that IT security leaders in Australia and New Zealand anticipate AI agents will address at least one of their organisation's digital security issues.

The survey reveals that all respondents see a role for AI agents in assisting with IT security, with 36 per cent of IT security teams in the region currently using such agents in their daily operations.

The proportion of security teams using AI agents is expected to grow rapidly, with predictions it will reach 68 per cent within the next two years.

According to the findings, 71 per cent of organisations in Australia and New Zealand are planning to increase their security budgets during the year ahead, just below the global average of 75 per cent.

AI agents were highlighted as being capable of supporting various tasks, including faster threat detection, more efficient investigations, and comprehensive auditing of AI model performance.

The global survey, which included more than 2,000 enterprise IT security leaders—with 100 respondents from Australia and New Zealand—also pointed to several challenges associated with adopting AI in security practices.

Despite widespread recognition that practices need to evolve, with 75 per cent of respondents acknowledging the need for transformation, 58 per cent expressed concern that their organisations' data infrastructure was not yet ready to maximise the potential of AI agents.

"Trusted AI agents are built on trusted data," said Alice Steinglass, EVP & GM, Salesforce Platform, Integration, and Automation. "IT security teams that prioritise data governance will be able to augment their security capabilities with agents while protecting data and staying compliant."

The report noted that while both IT professionals and malicious actors are integrating AI into their operations, autonomous AI agents offer an opportunity for security teams to reduce manual workloads and focus on more complex challenges. However, deploying agentic AI successfully requires a strong foundation in data infrastructure and governance.

In addition to familiar threats such as cloud security vulnerabilities, malware, and phishing, the report found that IT leaders now also rank data poisoning within their top three concerns. Data poisoning involves the manipulation of AI training data sets by malicious actors. This concern is cited alongside cloud security threats and insider or internal threats.