J.P. Morgan warns AI shrinks vulnerability window

J.P. Morgan Asset & Wealth Management published a report warning that artificial intelligence is dramatically reshaping the cybersecurity landscape by enabling faster discovery and exploitation of software vulnerabilities. The report states that the average time between vulnerability disclosure and exploitation has fallen to a single day, leaving companies little time to respond. AI models such as Mythos and GPT 5.5 are improving detection of unknown vulnerabilities, but the same capabilities can be exploited by malicious actors. The report notes that in about 60% of breaches, a patch was already available at the time of compromise, and highlights a global shortage of nearly 4.8 million cybersecurity professionals. It projects that the median time to exploit a vulnerability could decline to one minute by 2027. However, AI can also strengthen defenses, with security-focused tools being developed by Anthropic and OpenAI. The report urges businesses to prioritize rapid software updates and patch deployment.