Here’s what’s on our plate today:

• 🪖 How war drives AI innovation—and fuels tech worker revolts.

• 🗳️ Should tech workers have a say in how their AI is deployed?

• 💡 Roko’s Pro Tip helps you assess AI vendors for ethical red flags.

• 🧠 AI beats doctors, Meta limits kid bots, and Nvidia’s mystery clients drive Q2.

Let’s dive in. No floaties needed…

The Laboratory

How AI is reshaping militaries and big tech’s relationship with its employees

The relationship between war and technological progress is as old as human civilization. While the idea that many of today’s conveniences emerged from the machinery of conflict can be unsettling, history shows that military imperatives often accelerate innovation at speeds that cannot be matched solely through civilian research.

A great example of this can be found in the origins of the Internet. Created during the Cold War to ensure resilient communications, the U.S. Department of Defense’s ARPANET played a pivotal role in turning an experiment into a worldwide phenomenon. Regardless of whether the technologies are benign, like antibiotics, or capable of devastating the planet like nuclear weapons, wars inevitably fast-track the development and deployment of advanced tech. In this aspect, AI is no exception. The technology’s use in warfare is actively being debated, and not always for the right reasons.

Recently, Microsoft fired four employees who participated in protests on company premises against the firm's ties to Israel as it wages war in Gaza. The firings followed publication of an investigation by The Guardian, which revealed that Israel had used Microsoft cloud for expansive surveillance of Palestinians.

Earlier in 2024, Google fired 28 employees after some staff participated in a protest against the company's cloud contract with the Israeli government. These instances are part of the militarization of cloud infrastructure and AI services and reflect a shift in the role tech companies play in conflicts around the world. It also underscores the debate on how companies developing AI now play an important role on and off the battlefield and how their employees feel about it.

Where is AI being used?

Speed, scale, and pattern recognition are the core strengths of AI, and all of these, when combined, make the technology an ideal enhancement in modern warfare.

AI can process sensor data, satellite imagery, and battlefield communications in real time, providing commanders with actionable insights far faster than human analysts. And its use is no longer theoretical.

In May 2024, the U.S. awarded Palantir a $480 million prototype contract for the Maven Smart System. It uses data from various intelligence, surveillance, and reconnaissance sources, which are then fed to a machine-learning model. The model’s ability to learn from labeled examples is used to scan and identify enemy systems in the area.

Even before AI had captured the public’s imagination, the first use of autonomous robots was reported from Libya in 2021. A 2021 U.N. report described the STM Kargu-2 as capable of ‘fire-and-forget’ engagement without a data link.

Around the same time in 2021, Israel, according to a report from The Guardian, fought its ‘first AI war’ using machine learning and advanced computing. Since then, the country has reportedly deployed an AI target-creation platform called the Gospel, which, through the rapid and automatic extraction of intelligence, produced targeting recommendations for the IDF.

And when the technology is not in active deployment, war fronts are being used as sources of data to train future AI models.

Ukrainian officials have called the trove of information collected from the battlefields one of Kyiv's ‘cards’. The motive behind collecting the data sourced through millions of hours of combat footage collected via drones is to use it as a bargaining chip for more support from friendly nations.

Though data from the battlefield has historically been collected and saved by nation-states, it is now playing an important role in shaping relations between countries and the companies that use this data to develop more sophisticated systems.

How are tech companies involved?

Big tech companies like Microsoft and Google, though not directly linked to the development of weapon systems used on battlefields, provide support through cloud storage and underlying AI infrastructure.

According to an Associated Press report, Microsoft, in mid-2025, acknowledged that it sold advanced artificial intelligence and cloud computing services to the Israeli military during the war in Gaza and aided in efforts to locate and rescue Israeli hostages. At the time, the company also clarified that it had found no evidence that its Azure platform and AI technologies were used to target or harm people in the besieged enclave.

The acknowledgment came after an earlier investigation from AP revealed that the Israeli military was using Azure to transcribe, translate, and process intelligence gathered through mass surveillance, which can then be cross-checked with Israel’s in-house AI-enabled targeting systems and vice versa.

Google has also faced the ire of its workers for Project Nimbus, a $1.2 billion contract jointly awarded to Google and Amazon to supply the Israeli government with cloud services. While Google maintains that the Nimbus contract "is not directed at highly sensitive, classified, or military workloads relevant to weapons or intelligence services”, it has had to fire employees over protests.

Protests, terminations, and AI

Protests at big tech companies are not new. In 2018, Google was forced to shelve a contract with the U.S. military after a successful workers' protest.

However, with the AI boom, companies may not be open to giving up lucrative contracts, especially considering the massive infrastructural cost of building AI systems. Modern AI systems require massive data centers, which in turn need not only large swaths of land but also other resources like water for cooling the systems and electricity to power them. All of these require government support, which can be acquired more easily if defense contracts are involved.

In this scenario, the handling of employee protests by companies like Microsoft and Google risks setting a bad precedent for future disagreements in how and where technologies are deployed. They also underscore the importance of managing employee conscience when it comes to a transformative tech like AI in defense.

In their defense, Microsoft said its Azure and AI services for Israel’s Defense Ministry comply with terms and found no evidence they were used to target or harm civilians, following internal and external reviews. Google points to its AI Principles and public-sector cloud commitments, including Project Nimbus, to frame permissible workloads and safeguards.

AI’s rendezvous with the battlefield

The uneasy entanglement of artificial intelligence, militaries, and big technology companies is not new; it is a continuation of a centuries-old dynamic where war accelerates innovation.

In the 21st century, AI now stands at that same threshold as the Internet, GPS, antibiotics, and even nuclear energy have in the past. Its pattern-recognition capabilities, speed, and scale make it an unmatched tool for defense establishments, yet its diffusion into battlefields raises questions that extend far beyond the technical.

Unlike past eras, where military laboratories and government agencies drove much of the research, today’s breakthroughs in AI infrastructure are concentrated within corporations like Microsoft, Google, Amazon, and Palantir. These companies do not only supply governments; they also employ tens of thousands of workers, many of whom resist seeing their labor tied to war. That tension, manifested in protests, firings, and internal campaigns, has become as much a story of AI’s military adoption as the technology itself. The future, then, will hinge on how democratic societies, their militaries, and their tech industries create governance frameworks that both safeguard conscience and confront the realities of security.

Roko Pro Tip

Prompt Of The Day

Bite-Sized Brains

• AI Doctors Deserve a Second Opinion: A growing body of research suggests that AI can outperform doctors at detecting disease, but experts say the tech should assist, not replace, human care.

• Meta Sets Guardrails for Kids’ Chatbots: Meta released new guidelines for how developers should design AI assistants for minors.

• Nvidia’s Mysterious Megaclients: In a striking earnings call, Nvidia revealed that just two unnamed customers accounted for 39% of Q2 revenue.

Tuesday Poll

Rate This Edition