The new AI panic: dire warnings, missing guardrails, and a power struggle over who controls the future

AI isn’t just a tech story anymore—it’s colliding with politics, national security, media, and global power. In the last few months, warnings about AI have gone from niche expert debates to full-blown political talking points, with billionaires, former presidents, and lawmakers all trying to shape what happens next.

Anthropic’s warning: could AI systems “swarm” the internet?

A key flashpoint in the current AI panic comes from Anthropic, the company behind the Claude AI models. Its CEO has raised an especially stark concern: that AI systems are already helping to build the next generation of AI—and that this feedback loop could get dangerous fast.

In a recent article, he warns that within 6–12 months, a coordinated “rogue swarm” of AI agents could, in theory, spread across the internet, forming a persistent botnet capable of causing hundreds of billions of dollars in damage. The idea is that once AI is powerful and autonomous enough, it might be able to:

• Exploit software vulnerabilities at scale
• Hijack servers and devices to build massive botnets
• Disrupt financial systems, infrastructure, or communications
• Adapt faster than human defenders can respond

The warning is conditional—it depends on AI systems continuing to gain capabilities without matching progress on safety and control. But the core message is blunt: if powerful AI is deployed carelessly, the damage won’t be theoretical.

Elon Musk, other leaders, and the “whoever wins AI wins” mindset

What makes these warnings land harder is that other high-profile figures, including Elon Musk, have echoed many of the same fears about advanced AI. At the same time, there’s a competing narrative among political leaders that AI is a race—and whoever “wins” will dominate the future.

Donald Trump has leaned into this framing, arguing that the U.S. must stay ahead of China in AI at all costs. From this perspective, slowing down AI development or imposing strong regulations is framed as a risk to national power. The message is simple: “whoever wins AI wins.”

Critics push back on that logic. They argue that “winning” AI without strong safeguards may simply mean being first to unleash a technology that can’t be reliably controlled—raising the possibility that the supposed winner is also the first to be harmed by it. This tension—between speed and safety—is now at the center of global AI politics.

If you want a deeper dive into how AI leaders think about worst-case scenarios, including extinction-level risks, see this breakdown of Nate Soares’ stark AI warning.

Can we trust Big Tech to regulate itself?

There’s a growing unease about relying on the very companies building powerful AI systems to define the rules for how they’re used. When CEOs call for regulation, some observers hear a genuine plea for safety. Others hear something more self-interested: an attempt to shape rules that lock in their dominance and raise the barrier to entry for smaller competitors.

This skepticism shows up in several ways:

• Tech leaders warn of catastrophic AI risks—but still race to release more capable models.
• Calls for “guardrails” often lack specifics on enforcement, liability, or transparency.
• The same companies that talk about safety also profit from rapid deployment into sensitive areas like defense, finance, and critical infrastructure.

So we end up in a strange place: many people agree AI needs guardrails, but they don’t fully trust either the tech giants or the current political system to design them.

Congress admits it’s not ready—but wants “partnership” with industry

On the government side, there’s a growing recognition that lawmakers are behind the curve. The Speaker of the House has openly admitted that Congress is “less qualified” than AI companies to understand the technology and its implications. His proposed solution: a “partnership” between government and industry to define guardrails.

Critics see a problem here. If Congress feels unqualified and leans heavily on the same corporations that stand to profit from AI, there’s a real risk that:

• Regulations will be watered down or shaped to favor incumbents
• Strong, enforceable safety standards will be delayed or avoided
• Public interest—privacy, worker impact, democratic stability—takes a back seat to corporate goals

The result is a regulatory vacuum at the exact moment AI systems are being embedded into everything from cloud infrastructure to military planning.

Obama, Sanders, and the push to make AI oversight a core political issue

Not all political voices are treating AI as a side topic. Barack Obama has reportedly urged Democratic leaders to move AI oversight to the center of the party’s agenda, warning that the technology could become dangerous if they don’t get ahead of it. Bernie Sanders and other economic populists have been raising alarms as well, especially around how AI could impact workers and concentrate power.

Strategically, this reflects a shift: AI isn’t just about innovation and growth; it’s becoming a core voter issue tied to:

• Job security and automation
• Data privacy and surveillance
• Misinformation, deepfakes, and election integrity
• Environmental and local impacts of massive data centers

In many communities, residents are already pushing back against large AI data centers over concerns like energy use, water consumption, noise, and land impact. That grassroots resistance is starting to collide with bipartisan enthusiasm in Washington for big AI investments.

For a broader look at how AI policy debates are playing out globally—including religious and ethical perspectives—see our coverage of Trump’s AI executive order delay and the Pope’s call for human-centered AI.

Tech oligarchs, “billionaire psychosis,” and the fear of a “nerd Reich”

Some of the sharpest criticism isn’t just about specific AI tools but about the people and ideologies driving them. Writers and researchers who track Silicon Valley power politics argue that a subset of ultra-wealthy tech figures have moved into what’s been called “billionaire psychosis”: so much money and influence that they start to see the world as a sandbox for their own experiments.

In this view, AI is not just a product—it’s a lever to reshape society. Concerns include:

• Efforts to weaken democratic institutions in favor of technocratic or authoritarian governance
• Visions of a world where a small elite controls advanced AI and automation, while the rest of humanity is managed with basic income and digital systems
• Direct political projects and think tanks funded by tech billionaires to normalize these ideas

Authors like Gil Duran, in works such as “The Nerd Reich,” argue that some of these actors explicitly dream of a world government or post-democratic order run through technology, not public debate. Whether you agree or not, it’s a reminder that AI isn’t being developed in a political vacuum.

AI, the military, and media: where the lines blur

Another worrying trend is how tightly AI, the military, and major media outlets are starting to intertwine. Over the past few years, tech executives and AI contractors have been embedded in defense and national security institutions, helping to shape everything from battlefield analytics to logistics and intelligence.

At the same time, there are growing accusations that some governments are leaning on media partners to push favorable narratives about military operations and technology use. When defense officials allegedly pressure service members to appear on high-profile news programs to tell a specific story, it blurs the line between journalism and state messaging.

Combine this with AI-generated content, deepfakes, and automated influence campaigns, and you get a media environment where it’s harder and harder for the public to know what’s authentic, what’s spin, and what’s been quietly shaped by AI tools operating in the background.

Global power struggles: AI, China, Russia, and Ukraine

AI is also becoming a bargaining chip in broader geopolitical conflicts. The U.S.–China rivalry is often framed as a race for AI supremacy, with each side investing heavily in chips, models, and AI-enabled defense systems. That competition is used domestically to justify both acceleration and secrecy.

Meanwhile, the war in Ukraine has highlighted another dimension: how AI, data, and digital infrastructure intersect with diplomacy and conflict. Private envoys and politically connected figures have reportedly been involved in back-channel talks with Russian interests, raising questions about:

• Whose interests are really being represented in negotiations
• How much commercial and technological leverage is in play
• Whether AI and data deals are quietly bundled into geopolitical arrangements

When powerful but unelected actors shape foreign policy—especially in conflicts where AI-enhanced weapons, surveillance, and cyber operations are at stake—it adds another layer of opacity to an already complex situation.

Public distrust, midterms, and what voters can actually do

Underneath all of this is a widening trust gap. Many people don’t fully trust tech companies. They don’t fully trust governments either. And they’re understandably wary of AI systems that are rapidly getting more capable while oversight lags behind.

That’s why elections—especially midterms and down-ballot races—are becoming a key battleground for AI policy. Local and national officials will be the ones deciding:

• How aggressively to regulate AI companies
• What transparency and accountability standards to impose
• Whether to approve or block AI data centers and infrastructure in specific communities
• How to handle AI in policing, welfare systems, healthcare, and education

Some independent and progressive candidates are already running on platforms that treat AI as part of a broader economic and democratic struggle—tying it to labor rights, antitrust enforcement, climate policy, and civil liberties. On the other side, there are powerful incentives for both major parties to stay close to AI money and lobbying.

Where this leaves us: between panic and paralysis

The current AI conversation can feel overwhelming: rogue swarms, superintelligence, billionaire ideologues, geopolitical brinkmanship, and governments that admit they don’t fully understand what they’re regulating. But paralysis is exactly what benefits the status quo.

Moving forward, a few ideas are gaining traction across very different political camps:

• Independent oversight: Not just self-regulation by tech, but external bodies with real teeth and transparency.
• Slower deployment in high-risk areas: Especially in military, nuclear, bio, and critical infrastructure domains.
• Democratic input: Treating AI policy as a core civic issue, not a niche technical topic.
• Clear liability: Making it obvious who is responsible when AI systems cause harm.

AI isn’t going away. The question is whether its future is shaped primarily by a small cluster of tech oligarchs and political power brokers—or by a public that insists on safety, accountability, and a say in how this technology is used.