Cybersecurity: Outdated Software Fuels Surge in Digital Threats
New data shows a doubling of disclosed vulnerabilities this year, with AI tools accelerating the exploitation of legacy systems in critical infrastructure.
By Grant Sullivan ·

Cybersecurity experts warn that the United States faces growing digital threats due to the prevalence of outdated software in critical infrastructure. Aaron Mitchell, the chief executive of the cybersecurity firm HeroDevs, stated that attackers view legacy code as a primary target. He noted that these systems offer an easy entry point for bad actors to gain access and cause damage. Mitchell emphasized that the opportunity for exploitation is vast because of the sheer volume of old software still in use across various sectors.
The risk has intensified as artificial intelligence tools identify vulnerabilities at an accelerated rate. Last week, VulnCheck, a cybersecurity consulting firm, reported that more than 100,000 vulnerabilities have been found and disclosed this year. That number is more than double the count recorded in 2025. The firm also found that hackers are exploiting these flaws 33% faster compared to the previous year. This rapid pace leaves organizations with less time to implement fixes before an attack occurs.
Old software remains in operation across utilities, hospitals, and airlines because companies often lack the urgency or resources to replace it. Mitchell said that about half of all known exploits target versions that are no longer supported by the original maintainer. He described this proportion as rapidly growing. HeroDevs assists clients by maintaining support for open-source software that has reached end-of-life. The firm patches vulnerabilities in these systems to keep them secure after the original developers have moved on.
Recent incidents illustrate the tangible impact of these digital weaknesses. Earlier this month, grocery stores and butcher shops across seven European countries suffered a refrigeration hack. Hackers found a vulnerability in old software and raised the temperature of fridges and freezers to more than 100 degrees Fahrenheit. The attack was not particularly sophisticated but demonstrated how a long-standing flaw could disrupt operations across multiple nations. French blood banks also experienced a cyberattack earlier this month. Attackers attempted to increase temperatures using a similar tactic, though French authorities detected and stopped the intrusion.
Programming language flaws have also prompted widespread patching efforts. Last month, AI tools discovered two flaws in PHP, a language used to build more than two-thirds of websites. These flaws could allow an attacker to gain total access to a web server without a password. Mitchell described the response as a massive effort, noting that the internet was on fire for a week as developers downloaded patches. He confirmed that some websites were exploited before they could be secured. On Tuesday, an AI company revealed a critical vulnerability in Java known as Java4Shell, prompting companies to take immediate action to patch their systems.
Richard Horne, the head of the UK's National Cyber Security Centre, described these vulnerabilities as having been tolerated for far too long. The UK predicts that AI-driven cyberattacks on critical infrastructure will become a defining challenge by 2028. Mitchell believes that current exploitation rates are not yet at their peak. He warned that every day without remediation moves the industry closer to a tipping point where attacks become more frequent and damaging.
The current environment for digital security is further complicated by leadership changes in the United States. Axios reported on Thursday that Sean Cairncross, President Trump's National Cyber Director, will be leaving his role. Cairncross decided against serving an additional two-year term. His departure leaves a key position in the administration's cyber defense strategy vacant as the threat landscape continues to evolve.






