Big Tech profits from AI extinction hype, Ken Klippenstein says

In the debate surrounding the dangers posed by artificial intelligence, almost everyone is attempting to market a product.

While the mainstream media portrays artificial intelligence as an imminent mass extinction event, US President Donald Trump frames the issue within the context of a new Cold War with China, arguing that the US cannot afford to slow down.

In his analysis, journalist Ken Klippenstein emphasizes that Silicon Valley elites profit directly from this intense attention.

A new generation of the tech class is occupying the public mind with various doomsday scenarios, ranging from bioterrorism to machines taking over the world.

While this dynamic transforms artificial intelligence into a “national security” issue, it elevates the technology to the level of nuclear weapons and removes it from public oversight.

The intelligence community thinks the A.I. apocalypse narrative is BShttps://t.co/DpyAIXx4ob

— Ken Klippenstein (@kenklippenstein) September 15, 2026

Although not immune to the tendency to exaggerate threats, reports from US intelligence agencies paint a picture far removed from the hysterical tone in the media.

The US intelligence community assesses that artificial intelligence merely magnifies risks that already exist.

Sensational headlines run by legacy media are fueled by social media figures who spread claims of human extinction to millions of followers. In this way, fears themed around “existential risk” or “doomsday” take root in the public imagination.

On the other side of the coin are those who oppose disaster narratives while pursuing their own commercial interests.

Prominent figures in this camp include Yann LeCun, former chief AI scientist at Meta, and Andrew Ng, co-founder of Google Brain and head of AI Fund.

In October 2023, LeCun accused OpenAI chief Sam Altman, Google DeepMind chief Demis Hassabis, and Anthropic chief Dario Amodei of running a “massive lobbying effort” designed to tilt the regulatory landscape in their own favor.

LeCun warned that if these fear politics succeed, artificial intelligence will be monopolized by a small number of corporations.

Ng, for his part, described the claim that artificial intelligence would destroy humanity as “mind-bogglingly stupid”, arguing that large corporations are stoking extinction fears to avoid competing with open-source models.

Although these criticisms carry truth regarding corporate aims, the conflicting interests of both sides remain striking.

LeCun and Ng advocate open-source artificial intelligence models, whereas Anthropic and OpenAI favor proprietary models that keep their source code secret and lease access to users.

While major players producing proprietary models possess the capacity to comply with prospective federal licensing rules, open-source enterprises stand to be damaged by such statutory mandates. The common ground shared by doom-mongers and deregulation advocates is their lack of concern for the actual risks artificial intelligence generates.

The US intelligence community provides a more measured framework regarding tangible dangers.

In the Annual Threat Assessment, which catalogues China’s military strength, Russian influence operations, and drug cartels, the threats posed by artificial intelligence are summarized in just three items:

“It is essential to ensure that the use of machines and AI remains under human control.”

“These applications also carry risks that require careful human engineering to properly mitigate the risk of AI autonomy before they are widely deployed.”

“Emerging technologies such as AI and quantum computing are expected to have significant implications for national security.”

The official assessment by the 18 agencies comprising US intelligence on the perils of artificial intelligence remains limited to these statements. The reports contain no determinations concerning superintelligence, the annihilation of humanity, or an uprising of machines.

The US Department of Homeland Security Threat Assessment notes that artificial intelligence merely introduces fresh layers of complexity to existing threats. The department outlines five primary issues:

Disinformation, fabricated video or audio recordings (deepfakes), and election interference;

Cyber operations and financial crime;

The exploitation of this technology by violent extremists, alongside radicalization;

The proliferation of chemical and biological knowledge;

The circumvention of AI security controls and the poisoning of training data.

All of these risks were familiar prior to the emergence of artificial intelligence. AI-enabled disinformation permits legacy propaganda methods to be deployed with greater speed and intensity.

Cyber operations, financial crime, and elements of radicalization have likewise ranked as familiar subjects for many years.

Concerns regarding biological and chemical threats date back to the Bill Clinton administration, while the fifth item stems entirely from humans tampering with AI security controls.

The Global Catastrophic Risks Assessment report by the Pentagon-funded RAND Corporation think tank defines artificial intelligence as an “entropy source”.

The report states:

“AI can be thought of as adding entropy and chaos to thorny problems humans face. Chaos does not require the development of superintelligent or supercapable AI; it is possible with current and near-term AI capabilities.”

Another research study conducted within RAND identifies genetically engineered pathogens, geoengineering, and nuclear war as three plausible pathways to human extinction.

However, numerous physical and operational constraints prevent artificial intelligence from triggering these catastrophes.

In a follow-up report investigating whether large language models facilitate the planning of a mass biological attack, RAND Corp. researchers identified no statistically significant difference between plans formulated with AI assistance and those produced independently.

The US National Academy of Sciences similarly notes that the primary barrier to bioterrorism is not an absence of access to information.

The decisive bottleneck lies in DNA synthesis screening, hands-on laboratory skill, culturing, formulation, and aerosolization: procedures that all demand human intervention, carry high costs, and remain prone to failure.

Having no commercial product to release or corporate shares to protect, official analysts record with balanced language that artificial intelligence does nothing beyond compounding the scale and velocity of current problems.

Observing that all factions resort to exaggerated rhetoric to capture attention, Klippenstein points out that in the debate over whether a chatbot will transform into a god, the major actors turn a profit while foisting the cost onto the public.