Opinion: Hyperscalers head to the dark side

  • War! What is it good for? (Hooah!) Improving residual operating income after capital charges (Say it again!)
  • Big Tech’s unchecked Tolkien-themed militarization of benign technology is reshaping both the battlefield and the civil rights landscape
  • Policy area networks (PANs) present the comms industry with new challenges, beyond profitability and reliability 

The communications industry and the military have always shared technology, but until now, this has been one-way traffic. GPS, sensors, RFID, dual-ring optical network redundancy, and, of course, the internet (DARPAnet) were all developed for military use before being repurposed for civilian applications.

No longer. The polarity has now been inverted: so-called “dual use” technologies like AI, automation, and 5G, which were developed for peaceful purposes, are now being deployed as a keystone of 21st-century military strategy.

And this marks a fundamental shift in the nature of connectivity: away from the passive servicing of local- and wide-area network traffic, and toward the active enforcement of political mantra. Welcome, then, to the era of policy area networks (PANs).

This is big business. At MWC25 in Barcelona, Ericsson and Nokia, the twin stalwarts of the Western telecom industry, both pivoted toward defense as a core growth market, explicitly positioning their portfolios to tap expanding NATO and allied defense budgets now running into the hundreds of billions of dollars. The strategy helps both companies ameliorate the impact of weakening spending by carriers on their traditional product lines – telecom software, hardware, and services.

In September last year, I visited Nokia’s new R&D centre in Oulu, Finland, where I saw the company's first-ever telecom product – a radio made to order for the Finnish army in the 1960’s.

It seemed appropriate to ask its CEO, Justin Hotard, how the company would prevent its latest advanced 5G and AI technology from being sold to governments that might use it for unethical purposes. He told me that Nokia has created a formalized process of human rights due diligence and internal review committees. Ericsson has done the same, employing an exhaustive, multi-layered ethics, compliance, and export-control framework.

This is not really surprising. Both companies have supplied networking equipment to morally questionable organizations and administrations in the past, but have recognized the increased risks posed by new technologies such as AI and have taken steps to ensure they don’t repeat the same mistake. 

The concern is that a whole tranche of much larger organizations is now operating at the other end of the ethical spectrum: U.S. hyperscalers.

Big tech, bad people

Amazon (AWS), Microsoft, and Google Cloud supply technology to Ukraine for its war with Russia, bolstering the defense of a sovereign state under attack.

But those same hyperscalers have also embedded themselves as for-profit partners in Israel’s military cloud stack during the Gaza conflict, which UNICEF reports has resulted in the deaths of more than 17,000 Palestinian children and injuries to over 33,000 more. Under international law, the execution of indiscriminate or disproportionate attacks that foreseeably kill children constitutes a war crime.

According to investigative reports, Microsoft CEO Satya Nadella personally brokered access to a customized, segregated Azure environment for Unit 8200, the IDF’s signals-intelligence and cyber unit responsible for mass surveillance and data interception, which feeds directly into Israel’s military targeting and population control systems (Microsoft subsequently disabled access).

Beyond Microsoft, AWS and Google Cloud provide cloud and AI infrastructure under Project Nimbus, a $1.2 billion, multi-year contract supplying Israeli government and defense agencies—embedding their commercial hyperscale platforms deep within the operational machinery of modern warfare.

And then there’s Palantir, America’s Big Tech nightmare. Originally funded by the CIA’s venture arm, it operates on the belief that technology should concentrate data and decision power to enable states and institutions to dictate outcomes through the use of force — including facilitating coercion and violence —regardless of the price paid in transparency, consent, or civil liberties. Palantir’s code is at the epicenter of both the Gaza conflict and the current ICE mayhem in the U.S.

Out of bounds

While its attention-seeking CEO grabs most of the headlines, another of Palantir’s founders, Peter Thiel, is primarily responsible for defining its repressive mission.

Thiel personifies Big Tech’s loss of moral bearings. Privileged enough to have studied under the French philosopher Girard. Smart enough to grasp his warning that societies fueled by rivalry become disposed to normalizing violence against the weakest. And yet so profoundly lacking in self-awareness that he fails to see that, in Palantir, he has fabricated an industrial-scale accelerator of exactly the dynamics Girard warned about: amplifying imitative rivalry, legitimizing coercion, and operationalizing the very violence its creator claims to understand.

In practice, there are now few guardrails in place preventing hyperscalers run by executives who disregard basic moral axioms—like the idea that using their technology to help kill children is indefensible—from selling cloud and AI systems to governments committing war crimes or undermining democratic institutions.

Absent external enforcement, hyperscalers’ own staff have, in some instances, remonstrated in favor of their employers enforcing ethical use restrictions more rigorously. (Google’s response was to quickly fire dozens of the protesters). Shareholder activism, conversely, has been sparse and ineffective.

Selling off the Fourth Estate

The press’s role in providing independent oversight of corporate probity has been greatly reduced by the advent of the internet. What’s left of the trade journalist cadre reports on sanitized white-collar enterprise and consumer news, at a digital remove, with little or no attention to ethics.

Publishers are not complaining about the lack of coverage of the moral inversion of parts of the comms industry by their writers. I spent several months last year writing a report investigating hyperscalers’ involvement in unethical activities, including the sale of technology to countries accused of war crimes, tax evasion, and IP theft. It proved unprintable on either side of the Atlantic. Not because it was inaccurate. Publishers’ concerns were primarily commercial – hyperscalers now control the bulk of global advertising spend.

Another factor: the current U.S. Administration’s willingness to quash the distribution of inconvenient truths through coercion and legal action. Hyperscalers and government now operate in lockstep/goose-step when it comes to tech policy and deregulation. Eisenhower’s military-industrial complex has morphed into Trump’s military-hyperscaler complex, backed by the full weight of POTUS. This is not a coalition that B2B publishing companies want to get on the wrong side of.

Leave Frodo out of it

Companies driving the militarization of consumer technologies now sit at the very forefront of Silicon Valley’s entitlement and narcissism—rich in the same “smartest guys in the room” arrogance that once powered Enron straight into a deep, smoking crater.

That mindset is nowhere more evident than in their branding. Palantir, AndurilMithril… these firms reached for the lore of The Lord of the Rings to name their companies, as if Tolkien’s world were a neutral aesthetic rather than a moral statement. J.R.R. Tolkien’s mythology was forged out of the industrial slaughter of the First World War, shaped by trench warfare, mass death, and the dehumanizing machinery of modern conflict. His stories are literal warnings about power, domination, and surveillance.

Tolkien would have despised the appropriation of his writings by American Tech Bro narcissists. Twisting his moral codex into a naming convention for blood-soaked software, violence-for-profit, and machine-assisted social control is a desecration – and a perfect symbol of the amoral perfidy that now passes for technology innovation amongst the worst of the Silicon Valley elites.

Ultimus actus

Absent traditional protections against unethical behavior, decision-makers among the communications industry’s customers can still exert influence by choosing to purchase products and services only from vendors that do not enable the wholesale slaughter of innocents or threaten democratic norms. Those decisions are far more likely to be made by carrier executives than by their counterparts in the enterprise community, where hyperscalers originated. Telecommunications has always been a more principled and ethically constrained universe than the land of the cubicle. Long may that culture prevail.

Steve Saunders is a British-born communications analyst, investor and digital media entrepreneur with a career spanning decades.


Opinion pieces from industry experts, analysts or our editorial staff do not necessarily represent the opinions of Fierce Network.