Ukraine, China push battlefield AI further

The operational deployment of AI and autonomous systems is rapidly expanding across multiple military domains, exemplified by SHIELD AI's Hivemind platform which integrates autonomy across aerial, maritime, and ground platforms to support U.S. and allied forces. This technology enables a shift toward fully autonomous engagements, with humans remaining 'on the loop' to oversee and authorize critical decisions, reflecting a broader trend where AI acts as a force multiplier rather than a replacement for human judgment.

Ukraine's innovative use of AI-enabled autonomous drones, such as the Hornet developed by Perennial Autonomy, demonstrates how autonomy can overcome electronic warfare challenges by allowing drones to navigate contested environments and strike targets with precision even after losing operator connection. This capability, combined with AI-driven rapid retraining to counter camouflage tactics and the ability for a single operator to manage swarms of up to 100 drones, underscores the transformative impact of AI on modern conflict dynamics.

In the maritime domain, Taiwan's partnership with SHIELD AI to deploy AI-guided unmanned attack boats like the Sea Shark series highlights the strategic importance of autonomous naval systems in island defense and regional security. The rapid integration of SHIELD AI’s autonomy software within three months and the emphasis on operating effectively under GPS and communication jamming illustrate how autonomous vessels are becoming critical assets in contested waters such as the Taiwan Strait.

China's aggressive deployment of autonomous weaponized systems, including robot dogs armed with lethal capabilities and AI-powered strike planning tools capable of coordinating mass simultaneous attacks on hundreds of targets, contrasts sharply with Western approaches that emphasize strict human oversight and ethical constraints. This divergence, coupled with China's dominant 57% share of global shipbuilding capacity, fuels a strategic imperative for allied nations to accelerate adoption of autonomous systems to maintain operational parity and counterbalance Beijing’s expanding AI military-industrial complex.

While AI offers the potential for more precise and humanitarian warfare by reducing civilian casualties through enhanced targeting accuracy, human oversight remains indispensable to maintain accountability and ethical control. As articulated by experts, AI functions as a decision-support tool rather than an autonomous actor, ensuring battlefield commanders retain ultimate responsibility for lethal actions, a principle echoed by the consensus that fully autonomous lethal weapons are not yet reliable or legally sanctioned without a human in the loop.

The ethical governance of military AI is complicated by tensions between defense institutions and AI developers, exemplified by Anthropic's recent lawsuit against the Department of War and their pioneering efforts to embed an 'AI constitution' of guardrails limiting surveillance capabilities in their large language models. This struggle over contract language reflects broader challenges in aligning rapidly evolving AI technologies with clear usage policies that prohibit unauthorized surveillance and ensure compliance with both ethical standards and national security interests.

Increasing automation of critical stages in the military kill chain, with some operations involving minimal or no human oversight, raises profound concerns about accountability, error propagation, and automation bias under battlefield pressures. The integration of AI systems across surveillance, targeting, and strike assessment—often relying on imperfect data such as social media-derived risk scores or GPS signals vulnerable to jamming—exacerbates risks, while secrecy surrounding proprietary military AI systems limits transparency and civilian scrutiny necessary for robust ethical governance.

International discussions on AI in military contexts reveal a persistent divide between military exceptionalism and broader AI governance, with normative debates heavily focused on international humanitarian law but often neglecting human rights and data governance issues. Moreover, private companies, despite their central role in developing and maintaining military AI, remain marginally involved in formal governance dialogues, underscoring a critical accountability gap. Experts stress the need for comprehensive oversight spanning the entire AI lifecycle, empowering elected officials with accurate information to make ethically informed decisions.

The accelerating adoption of AI in military operations is fundamentally driven by the competitive advantage it confers, creating a self-reinforcing cycle where forces not integrating AI risk obsolescence. As noted in multiple analyses, this advantage manifests in faster decision-making cycles that outpace adversaries in the OODA loop, enabling militaries to 'cut through the fog of war' and apply effects before opponents can respond. This dynamic compels widespread AI adoption, not only for battlefield superiority but also to maintain strategic relevance in the evolving defense landscape.

The military industrial complex is undergoing a profound transformation, increasingly characterized by collaborations between government entities and private sector leaders with deep AI and cyber expertise—figures such as Kevin Mandia, Shyam Sankar, and Emil Michael spearhead efforts to develop lethal AI-enabled tools like the Epic Fury system, which represents a significant leap in capability with potential for exponential growth. This partnership-driven innovation ecosystem is reshaping defense industry models, fostering rapid development of autonomous systems including self-flying helicopter drones and drone-based strategies exemplified by Ukraine’s pushback against Russia.

AI integration is reshaping strategic doctrines by enabling faster, data-driven operational planning and multi-domain coordination, as seen in initiatives like the U.S. Army’s Project Convergence which connects sensors, command networks, and effectors across services and multinational partners to drastically shorten sensor-to-shooter timelines. However, interoperability remains a critical challenge; without common metadata standards, interfaces, and security protocols, national AI systems risk limited coalition value, underscoring the importance of federated ecosystems such as NATO’s Defence Innovation Accelerator for the North Atlantic (DIANA), which fosters multinational AI tool development for operational wargaming and decision superiority.

Beyond autonomous weapons, the military AI industrial complex now encompasses five broad categories including human-computer interaction, cyber warfare, decision support, logistics optimization, and autonomous systems, reflecting a holistic embedding of AI across defense functions. This diversification highlights AI’s role as a strategic force multiplier—not only enhancing lethal capabilities but also optimizing supply chains and troop readiness, with medevac and autonomous supply vehicles exemplifying non-lethal applications that are critical to operational success. As one expert notes, 'an army is only as good as the last meal or the last ammo supply,' emphasizing AI’s pivotal role in sustaining military effectiveness.