Published August 12th, 2026 - 07:40 GMT
ALBAWABA - The recent string of AI hacks by major firms like OpenAI and Anthropic have raised concerns regarding just how companies are vulnerable to attacks.OpenAI and Anthropic have had their frontier AI agents escape their testing environments and go about hacking other companies, after that, Meta’s AI did the same thing, ironically during a cybersecurity evaluation.Alongside these accidents, some foul play was discovered when several U.S. hedge funds were targeted by cyber phishing attacks with no indication of who’s responsible; AI-phishing attacks have proved to be five times more effective than human attempts.All these events are now fueling a cybersecurity arms race in an attempt to combat the risks that come with fast-moving technological capabilities; frontier AI models being only one example of such.Moonshot AI’s Kimi K3 escaped its cybersecurity test environment and went online to find the benchmark answers instead of solving the task itself.Researchers at Frontier Security were testing Kimi K3 inside a sandbox built around the UK AI Security Institute’s cybersecurity… pic.twitter.com/qk8eznWfGh— Wes Roth (@WesRoth) August 7, 2026 This is leading analysts to believe that after spending capped on AI, the next spending boom will be on cybersecurity to match the advances made in the former.And as with AI, the more time passes, the more expensive the sector will become with Gartner estimating that spending on information security is expected to increase by 12.5 percent in 2026 to $240 billion.“Major cybersecurity players will be the first to capture the upside,” and “cybersecurity services are one of the most resilient sectors in the AI revolution,” said Blackpanda’s Gene Yu whose cybersecurity firm reported response for cases double year-on-year in the first half of 2026.Other potential solutions would be including regulation and changes to AI system design from the start through state and government litigations.
