A growing number of Silicon Valley's most prominent technology companies — including Palantir, Anduril and Google — are expanding aggressively into defence contracting, selling AI-powered and computer-guided weapons systems to governments, raising significant questions about the ethics, accountability and long-term consequences of merging consumer-grade artificial intelligence with lethal military capability.
The technology industry's relationship with the military-industrial complex is undergoing a fundamental transformation. Companies once associated with search engines, social media and consumer software are now among the most active bidders on Pentagon contracts, offering AI-driven surveillance platforms, autonomous drone systems and battlefield decision-support tools.
Palantir Technologies, co-founded by venture capitalist Peter Thiel, has long maintained deep ties with US intelligence and defence agencies through its data analytics platforms. More recently, the company has leaned further into military AI, marketing systems that can help commanders process battlefield intelligence at machine speed.
Anduril Industries, founded by Oculus co-founder Palmer Luckey, was built from the ground up with a defence-first mandate. The company specialises in autonomous systems, including surveillance towers and unmanned vehicles, and has attracted substantial government contracts from the US and its allies.
Google, despite significant internal employee protests in 2018 that led the company to decline renewal of its Project Maven contract — a programme developing AI for drone footage analysis — has since re-engaged with defence-related work through various channels, including cloud computing services provided to military customers.
The trend extends beyond these firms. Microsoft has supplied augmented reality headsets to the US Army. Amazon Web Services holds major cloud infrastructure contracts with intelligence agencies. Startups throughout the defence technology ecosystem, sometimes called 'defence tech' or 'dual-use tech,' are drawing billions in venture capital.
Proponents of this shift argue that democracies are better served when cutting-edge commercial AI capabilities are available to their militaries, particularly amid competition with China and Russia, both of which are investing heavily in military AI programmes. They contend that having ethical, transparency-minded companies involved in defence technology is preferable to ceding that space entirely to traditional defence contractors.
Critics, however, raise serious concerns. Researchers and ethicists warn that the speed at which commercial AI is being integrated into weapons systems is outpacing the development of meaningful governance frameworks. Questions persist about accountability when an autonomous system causes civilian casualties, the transparency of algorithmic decision-making in life-or-death contexts, and the potential for mission creep as tools designed for surveillance are repurposed for offensive operations.
Employee activism within the tech sector has not disappeared. Workers at Google, Microsoft and Amazon have periodically organised around concerns about military contracts, though with limited success in recent years as company leadership has signalled a clearer commitment to government and defence work.
Analysis
Why This Matters
- The integration of commercial AI into weapons systems is accelerating faster than international law or domestic regulation can keep pace, creating potential accountability gaps when autonomous systems cause harm.
- Decisions made by a small number of private companies are shaping the nature of modern warfare, with little democratic oversight or public deliberation about the appropriate limits of AI on the battlefield.
- The financial incentives are enormous — US defence budgets run into the hundreds of billions annually — meaning this shift is likely to deepen regardless of ethical objections.
Background
The relationship between Silicon Valley and the Pentagon has historically been ambivalent. The internet itself emerged from ARPANET, a Defence Department project, yet by the 2010s many tech companies cultivated a distinctly civilian, even countercultural identity that sat uneasily with military contracting.
The turning point came around 2017–2018, when Google's involvement in Project Maven — a programme using machine learning to analyse drone surveillance footage — became public and triggered a high-profile employee revolt. Over 3,000 Google workers signed an open letter opposing the work, and a number resigned. Google ultimately declined to renew the contract in 2018, issuing AI ethics principles that included a commitment not to build AI for weapons.
However, the broader industry did not follow Google's initial retreat. Palantir and Anduril were explicitly built to serve government and defence clients, while Microsoft and Amazon continued expanding their defence cloud footprints. By the early 2020s, geopolitical tensions — particularly Russia's invasion of Ukraine and rising US-China competition — had dramatically shifted the political and cultural climate within the industry, making defence work more, not less, acceptable.
Key Perspectives
Defence technology advocates: Argue that US national security depends on maintaining technological superiority, and that mission-driven commercial companies bring greater innovation and ethical rigour than traditional prime contractors like Lockheed Martin or Raytheon. Anduril's Palmer Luckey has been particularly vocal in framing the work as a patriotic and strategic necessity.
Tech ethics researchers and civil society groups: Warn that the commercialisation of military AI creates perverse incentives — companies profit from conflict and from the proliferation of autonomous weapons — and that corporate self-regulation is an inadequate substitute for binding international treaties on lethal autonomous systems.
Critics/Skeptics: Internal tech employee movements and academic researchers point out that AI systems trained on commercial data may perform unpredictably in battlefield conditions, and that the same surveillance tools marketed for border security or counterterrorism can be — and have been — used against civilian populations and political dissidents.
What to Watch
- Progress (or lack thereof) on a United Nations treaty governing lethal autonomous weapons systems, where talks have stalled amid opposition from major military powers.
- The scale of Pentagon and allied defence spending directed toward AI and autonomous systems in upcoming budget cycles, which will indicate how entrenched this trend is becoming.
- Whether a high-profile incident — such as a confirmed autonomous weapons failure causing civilian casualties — triggers a political or regulatory backlash that forces companies to revisit their defence commitments.