Artificial Intelligence Is Growing More Powerful — Who Is Watching the Watchers?
A Walk My Town Staff Opinion
Technology, Community & Public Affairs
Artificial intelligence is quickly becoming part of everyday American life. We use AI to research information, communicate, conduct business, analyze documents, create content and increasingly assist with decisions that previously required substantial human effort.
The potential benefits are extraordinary. AI may improve medicine, education, scientific research, transportation, government services and cybersecurity.
But as artificial intelligence becomes more capable and more autonomous, Walk My Town believes the public should also be asking a fundamental question: Who is watching the watchers?
The Greater Risk May Be How Humans Use AI
Public discussions about artificial intelligence sometimes focus on a dramatic possibility: Could AI eventually become so powerful that humans lose control of it?
That is an important subject for researchers to examine, but it should not distract us from risks that already deserve attention.
AI does not need to become conscious or independently “take over” society to cause harm. Criminals, hostile governments and other malicious actors can potentially use increasingly sophisticated AI tools to make fraud, impersonation, misinformation and cyberattacks more effective.
The National Security Agency has acknowledged this emerging challenge. Its Artificial Intelligence Security Center says AI offers significant opportunities but also creates a substantial new attack surface, and the NSA reports that American adversaries are already testing tools against this technological environment.¹
For ordinary citizens, that makes AI security much more than an abstract technology issue.
It is potentially a community and public-safety issue.
What Happens to Our Personal Information?
Consider how much of our lives now exists electronically: banking records, medical information, photographs, passwords, tax information, Social Security numbers and personal communications.
Artificial intelligence can be used to help protect this information. But AI capabilities can also make fraudulent communications and other forms of cybercrime more sophisticated.
The FBI advises Americans to be alert to AI-generated content, including deepfakes and synthetic audio. At the same time, the FBI itself uses AI for certain investigative purposes—but requires human experts to assess AI output before substantive action is taken.²
That principle deserves wider public discussion.
When artificial intelligence affects someone’s finances, healthcare, safety, employment or fundamental rights, should a human always remain responsible for the final consequential decision?
Walk My Town believes this is a question worth asking now, rather than after these technologies become embedded throughout society.
What About Our Critical Infrastructure?
The stakes become considerably greater when AI is connected to the systems upon which communities depend.
Electricity. Water. Hospitals. Telecommunications. Transportation. Financial services.
The Cybersecurity and Infrastructure Security Agency has warned that incorporating AI into critical-infrastructure systems may introduce vulnerabilities to critical failures, physical attacks and cyberattacks. CISA has called for identifying and mitigating these risks as AI adoption expands.³
The National Institute of Standards and Technology is also addressing this issue. In April 2026, NIST announced work on a specialized AI Risk Management Framework profile for trustworthy AI in critical infrastructure.⁴
This does not mean AI is about to independently shut down an electrical grid or water system.
It means the federal agencies responsible for technology standards and cybersecurity recognize that AI interacting with critical infrastructure requires serious risk management.
Humans Must Remain in Control
One of the most important principles identified by NIST is surprisingly straightforward.
AI systems should be monitored.
NIST’s AI Risk Management Framework discusses rigorous testing, real-time monitoring and the ability to shut down, modify or allow human intervention when an AI system deviates from its intended or expected operation.⁵
NIST’s broader framework also calls for clearly defined accountability and mechanisms to disengage or deactivate AI systems whose performance or outcomes become inconsistent with their intended use.⁶
These principles should not be controversial.
We require safeguards for automobiles.
We establish standards for aircraft.
We protect drinking-water systems.
We impose security requirements on banks and hospitals.
Why should increasingly autonomous artificial-intelligence systems be treated differently?
Who Should Be Accountable?
Perhaps one of the clearest principles comes from the FBI’s own policy regarding its use of artificial intelligence.
The Bureau says AI-generated information used in investigations must be reviewed by trained investigators or analysts. Its position is straightforward: **a human being remains ultimately accountable for the actions taken, not the AI.**⁷
Walk My Town believes this concept deserves consideration well beyond law enforcement.
As artificial intelligence becomes increasingly capable, society should determine where responsibility ultimately rests.
Is it the AI developer?
The company deploying the system?
The government agency using it?
The person supervising it?
And when several autonomous AI agents are communicating and completing tasks together, who is responsible for their collective actions?
Technology should never become an excuse for eliminating human accountability.
Innovation and Safety Can Coexist
Walk My Town does not believe society should respond to artificial intelligence primarily with fear.
AI has enormous potential to improve our lives.
It could help physicians recognize diseases, help scientists make discoveries, strengthen cybersecurity, improve education and make government and businesses more efficient.
But supporting technological innovation does not require abandoning reasonable safeguards.
NIST’s voluntary AI Risk Management Framework itself was developed to help organizations manage AI risks while enabling society to benefit from the technology.⁸
We believe that is the appropriate balance.
Innovate—but establish guardrails.
Encourage technological development—but protect privacy.
Use AI to improve cybersecurity—but protect critical infrastructure from AI-related vulnerabilities.
Allow automation—but maintain meaningful human oversight for consequential decisions.
And as autonomous systems become more powerful, preserve the ability for people to intervene, deactivate them and remain accountable for what they do.
A Question for Our Readers
Artificial intelligence may become one of the most transformative technologies of our lifetime.
The decisions being made today could influence how these systems operate in our communities for decades.
That makes this more than a conversation for Silicon Valley, Washington, D.C., computer scientists or technology companies.
It is a conversation for all of us.
How much authority should we give artificial intelligence?
What decisions should always require human approval?
Should autonomous AI ever have direct control over critical infrastructure?
Who should be legally responsible when an AI system causes significant harm?
And who should have the authority to shut an AI system down?
Walk My Town welcomes respectful public discussion of these questions.
Artificial intelligence can become an extraordinary tool for humanity.
Our challenge is making certain that humanity remains the one holding the tool.
⸻
Footnotes & Sources
¹ National Security Agency (NSA), Artificial Intelligence Security Center (AISC). The NSA established the AISC as part of its cybersecurity mission. The agency states that AI presents significant opportunities while opening a large and volatile attack surface and that U.S. adversaries are already testing tools against this emerging environment.
² Federal Bureau of Investigation (FBI), Artificial Intelligence — Emerging and Advanced Technology. The FBI discusses threats involving deepfakes and synthetic media as well as its own controlled uses of AI. The Bureau states that AI-generated investigative leads are verified and validated by human experts.
³ Cybersecurity and Infrastructure Security Agency (CISA), 2023–2024 CISA Roadmap for Artificial Intelligence. CISA states that deploying AI may make critical-infrastructure systems more vulnerable to critical failures, physical attacks and cyberattacks and calls for identifying and mitigating those vulnerabilities.
⁴ National Institute of Standards and Technology (NIST), AI Risk Management Framework. On April 7, 2026, NIST released a concept note for an AI Risk Management Framework Profile on Trustworthy AI in Critical Infrastructure, intended to guide operators toward risk-management practices when using AI-enabled capabilities.
⁵ NIST, Artificial Intelligence Risk Management Framework (AI RMF 1.0), Section 3.2 — Safe. NIST identifies practical AI-safety approaches including rigorous testing, real-time monitoring and the ability to shut down, modify or provide human intervention when systems depart from intended or expected functionality.
⁶ NIST, AI Risk Management Framework Core. NIST recommends defined organizational accountability, ongoing monitoring and mechanisms to supersede, disengage or deactivate AI systems demonstrating performance or outcomes inconsistent with their intended use.
⁷ Federal Bureau of Investigation, Artificial Intelligence — Emerging and Advanced Technology. In explaining its own AI safeguards, the FBI states that trained investigators or analysts must assess AI output before substantive action and that human beings ultimately remain accountable for actions taken.
⁸ NIST, Artificial Intelligence Risk Management Framework (AI RMF 1.0), January 26, 2023. The voluntary framework is intended to help organizations designing, developing, deploying or using AI manage risks while promoting trustworthy and responsible AI development and use.
⸻
Public Disclosure & Editorial Note
This article represents opinion and commentary by Walk My Town staff and is published for public review, education and discussion. Questions concerning future AI autonomy, human control, accountability and the appropriate limitations on artificial-intelligence systems are presented as matters for public discussion and should not be interpreted as claims that current AI systems have independently taken control of critical infrastructure or other institutions.
Factual statements attributed to U.S. government agencies are supported by the government publications identified in the footnotes. Readers are encouraged to review the original sources and reach their own conclusions.
Walk My Town — Technology, Community & Public Affairs
Community perspectives. Public issues. Respectful discussion.
