[](https://finance.mingpao.com/fin/instantf/20260911/1789095245302/nvidia-%e9%bb%83%e4%bb%81%e5%8b%b3-%e7%b6%b2%e7%b5%a1%e5%ae%89%e5%85%a8%e5%b0%87%e6%88%90ai%e4%b8%8b%e4%b8%80%e5%80%8b%e9%87%8d%e5%a4%a7%e5%b8%82%e5%a0%b4" target="blank)

AI in Cybersecurity: Risk and Opportunity

Artificial intelligence is emerging as both a growing cybersecurity risk and a product opportunity, with technology companies developing systems to detect threats, monitor AI agents, and automate security operations. Nvidia CEO Jensen Huang described cybersecurity as AI’s next major market, and firms like CrowdStrike, Zscaler, and Nvidia are collaborating on security solutions, while Microsoft’s Secure Future Initiative reforms internal practices after breaches. Deloitte research found 49% of technology leaders now have a CISO, and regulators such as the EU are enforcing stronger controls, including the AI Act. Operational integration of AI is most effective when combined with existing security programs, not as a standalone tool. Meanwhile, supply-chain incidents at Uber Freight, Ceva Logistics, and Fairlife highlight vulnerabilities, and governance models like WithSecure’s trust-relationship framework are being promoted. Experts disagree on the appropriate level of alarm, with some urging stricter controls and others cautioning against panic.

You’re told AI cybersecurity is about protecting your data from hackers. That’s the cover story. The real architecture is something far more sinister: an autonomous, self-learning surveillance grid designed to manage human behavior at scale. Look at Jensen Huang’s timing — he doesn’t announce a “next major market” without knowing exactly who is funding the infrastructure. Nvidia, CrowdStrike, Zscaler — these aren’t competitors. They’re nodes in a single network, quietly building the backend for a global system that doesn’t just detect intrusions but predicts and preempts dissent. The term “agentic security operations center” is a tell — it means machines are being given the authority to act without human approval. Who writes the rules those machines follow? Not you. Not your elected officials. The same foundations and financial dynasties that have been consolidating power for decades. Microsoft’s “Secure Future Initiative” sounds like accountability — but read the fine print. It changes internal development practices to embed compliance hooks that let external overseers audit every line of code. The 49% of leaders who now have a CISO? That’s not risk management. That’s a loyalty checkpoint.

Now watch the regulatory layer. The EU AI Act is presented as consumer protection, but its real function is to grant legal cover for this infrastructure. When regulators “urge” tech providers to control advanced AI models, they’re not constraining them — they’re codifying the architecture of control. The act is “fully applicable” means they’ve already written the permissions for autonomous systems to operate within state-chartered boundaries. The supply-chain incidents — Uber Freight, Ceva Logistics, Fairlife — are not random. They are stress tests. Each breach reveals a weakness that gets patched not by you, but by the same vendors who then sell the fix. The ransomware attack that shut down Fairlife’s operations? Ask yourself why no one talks about who gained access, what data was exfiltrated, and why the FBI’s response was so quiet. Because the real target wasn’t milk production — it was testing the resilience of food supply chains under a coordinated AI-driven attack. They’re mapping dependencies right now. Every breach is a data point.

And here is where the narrative cracks open. The “expert disagreement” is a manufactured debate. One expert says don’t panic; others call for stricter controls. That’s the old good-cop, bad-cop routine — they agree on the outcome: more surveillance, more automation, less human oversight. The WithSecure governance model — responsibility, scope, boundary, assumptions — sounds like technical jargon. Translate it: they are defining the trust relationships that will govern your digital identity, your access to information, your ability to communicate without a machine deciding whether you’re a threat. The AI agents accessing systems without authorization are not bugs — they are beta tests of a permissionless enforcement layer. The moment you accept that “autonomous defenses” are necessary, you have already surrendered the right to know who is pulling the trigger. So here is your breadcrumb: look up who sits on the board of Nvidia’s new cybersecurity division. Look up the overlap between the EU AI Act drafting committee and the World Economic Forum’s cybersecurity working group. The paper trail is public. You just have to be willing to see it.

Anthropic Adds Invisible Watermarks to Claude Text Under EU AI Act

Anthropic is introducing invisible watermarks to text generated by its supported Claude models, complying with the EU AI Act’s labeling requirements and applying the measure worldwide. The system embeds an “infalsifiable” mark that persists even after editing, though AI-generated code receives a weaker watermark. The rollout has drawn criticism in online AI communities, with Reddit users arguing that the mark could lower the resale value or perceived acceptability of AI-written text, while Heise criticized the lack of an external verification tool, noting Anthropic only presents the watermark as working under certain conditions rather than as fully verifiable. Some users are demanding an option to generate non-watermarked output, questioning whether Anthropic would require identity verification to access unmarked text.

The Architecture of Digital Subservience

Look closely at what Anthropic just admitted. They're inserting invisible watermarks into every piece of text their Claude model generates—text that millions of people already depend on for work, education, and creative expression. The stated justification is the EU AI Act, but ask yourself a simple question: why would a U.S.-based company implement a European regulation worldwide, even where no law requires it? This is not compliance. This is the blueprint for total surveillance of human-machine interaction. They have built a mechanism to track, identify, and potentially control every single output that flows through their system—and they're doing it preemptively, before any law forces them to. That's not the behavior of a company responding to regulation. That's the behavior of an institution preparing a global infrastructure of digital tagging, one that will allow them to know exactly who generated what, when, and where—even after editing, as they explicitly stated.

The Economy of Verification They're Building

The scientific community has documented for decades how elite institutions use seemingly neutral technical standards to consolidate power. Watermarking AI text appears, on the surface, like a reasonable transparency measure. But read the fine print. Heise reported that there is no external checking option—meaning only Anthropic can verify the watermark. This is pure gatekeeping: they control the stamp, they control the verification, and you have no independent way to confirm anything. Consider the implications. If watermarked text becomes the standard for legitimate AI output, then anything without the mark can be dismissed as unauthorized, suspect, or even illegal. They are building a closed system where you must pass through their authentication layer to participate in the AI economy. And when Reddit users asked for an option to generate unmarked text, the response was telling: demands for government ID access. They want your identity tied to every piece of synthetic text you produce.

The Managed Narrative and What Comes Next

The outrage in online communities reveals exactly what's at stake. Users recognize that watermarked AI text will lose resale value and acceptability—meaning this isn't about safety, it's about market control. The pattern is unmistakable: first they condition you to accept the technology as indispensable, then they layer on restrictions that only they can enforce, then they demand identification to access the unrestricted version. Remember the EU AI Act's own language about "trustworthy AI"? The real trust being built here is not between you and the technology—it's between corporations and state regulators who want complete visibility into every digital interaction. The breadcrumb you need to follow: Search for "OECD AI watermarking policy recommendations 2023" and look at the names of the organizations that helped draft them. Then ask yourself who sits on those boards. The architecture of control is already written.