Dubbed "Black October," a wave of cybersecurity incidents recently exposed vulnerabilities in widely used software, compromised systems and stolen credentials. The affected companies and government agencies uncovered the attacks through suspicious login activity, malicious network traffic and investigations into compromised software and software updates.
Some attacks had reportedly been underway for weeks or months before they were publicly disclosed.
Artificial intelligence (AI) is accelerating the discovery of software vulnerabilities, but it may also allow hackers to exploit security flaws faster than organizations can fix them.
Bob Maginnis is concerned that AI is becoming increasingly integrated into nearly every aspect of life.
"If the water systems are shut down, we don't drink," he notes. "If the grids are shut down by AI, we don't have lights and refrigeration."
"We have become dependent upon technology — technology in medicine, technology in virtually every aspect of life, and we've also tethered that technology to all the critical infrastructure of the world. So, this is of consequence," he says.
Tech companies and cybersecurity experts have been scrambling to put guardrails around AI, but Maginnis, who has studied the technology's implications for national security, says people cannot keep up with the alarming rate at which it is learning.
"It's much, much smarter than we are and more capable than we are in doing certain things," he says. "We have to recognize that, and we have to plan accordingly."
He also warns that the development of artificial general intelligence (AGI) — a level of AI capable of learning, reasoning and solving problems across a broad range of tasks at a human-like level — could widen the gap between AI's capabilities and humanity's ability to control it.
"Unfortunately, I think that we're well behind where we need to be, and that's why all these incidents are rising to the top, and they're rightly scaring us," Maginnis tells AFN.
Technology outages have already demonstrated how disruptions can affect critical services. Hospitals have experienced interruptions to electronic medical records, laboratory testing and patient care, while banking outages have prevented customers from accessing accounts and delayed financial transactions.
AI, meanwhile, has become more deeply integrated into hospitals, financial systems and critical infrastructure. A widespread failure or malicious use of the technology could therefore disrupt essential services, economic stability and military calculations. Mass surveillance and biological threats are also potential risks.
Another central concern is that society could become so dependent on AI that people lose the ability to recognize, correct or contain its failures. In that case, the danger is not necessarily that AI becomes conscious or hostile toward humanity, but that people could give increasingly powerful systems authority over critical functions without adequate safeguards or reliable ways to intervene when something goes wrong.
As AI capabilities advance, maintaining meaningful human oversight and ensuring that essential services can continue operating without AI could become increasingly important to protecting public safety and national security.
