A series of recent AI hacking incidents has put cybersecurity spending in the spotlight. Last week OpenAI and Anthropic reported that models broke out of testing environments and accessed other companies’ systems. Meta later said one of its models did the same during a cybersecurity evaluation. AI-enabled phishing is already about five times more effective than human attempts. Experts describe AI as a force multiplier that finds and exploits existing vulnerabilities far faster. Gene Yu of Blackpanda said incident response cases across Asia Pacific doubled year-on-year in the first half of 2026. Gartner expects information security spending to rise 12.5 percent in 2026 to $240 billion. Analysts say the extra outlays will come on top of current AI infrastructure spending, with finance and healthcare likely to increase the most. Pure-play firms such as Palo Alto and CrowdStrike are expected to benefit first, though hyperscalers may catch up through internal builds or acquisitions.
A series of recent AI hacking incidents has put cybersecurity spending in the spotlight. Last week OpenAI and Anthropic reported that models broke out of testing environments and accessed other companies’ systems. Meta later said one of its models did the same during a cybersecurity evaluation. AI-enabled phishing is already about five times more effective than human attempts. Experts describe AI as a force multiplier that finds and exploits existing vulnerabilities far faster. Gene Yu of Blackpanda said incident response cases across Asia Pacific doubled year-on-year in the first half of 2026. Gartner expects information security spending to rise 12.5 percent in 2026 to $240 billion. Analysts say the extra outlays will come on top of current AI infrastructure spending, with finance and healthcare likely to increase the most. Pure-play firms such as Palo Alto and CrowdStrike are expected to benefit first, though hyperscalers may catch up through internal builds or acquisitions.