AI Targeting Blind Spots and a Cyber Talent Push—Are Defense Systems Losing Control?

Two Small Wars Journal pieces published on 2026-10-01 argue that modern defense workflows are becoming less reliable as reliance on AI and automation grows. One article focuses on “AI over-trust” and claims this can degrade the joint targeting cycle by reducing human verification, slowing feedback loops, and amplifying model errors into operational decisions. The second piece, “The Force We Cannot Count,” frames a related risk: forces and capabilities may be harder to measure than planners assume, especially when systems, data, and readiness signals are fragmented across domains. Together, the articles suggest a structural mismatch between how militaries quantify effects and how AI-enabled systems actually perform in complex, small-scale conflict environments. Strategically, the core geopolitical implication is that deterrence and escalation management may weaken if militaries cannot reliably predict outcomes from AI-assisted targeting and force assessment. If AI confidence is treated as equivalent to operational certainty, adversaries can exploit uncertainty through deception, electronic warfare, or information operations that distort sensor inputs and degrade model performance. The “cannot count” framing also points to a governance and command-and-control challenge: when readiness and capability visibility deteriorate, decision-makers may either overreact to noisy signals or underreact to genuine threats. In this context, the likely winners are actors that can manipulate data quality and decision timelines, while the likely losers are forces that depend on centralized, automated assessments without robust validation. On markets, the direct commodity impact is not explicit in the provided articles, but the defense and cybersecurity angle implies second-order effects for defense contractors, cyber services, and secure communications providers. If AI targeting and joint fires processes face credibility gaps, procurement priorities could tilt toward verification, human-machine teaming, and resilient command-and-control—areas that typically support demand for defense software, simulation, and cyber hardening. The Australian defense cyber skills item from defence.gov.au adds a concrete signal that at least some governments are investing in specialized cyber talent, which can support domestic cyber ecosystems and contractors tied to incident response and defensive operations. For investors, the most relevant instruments would be defense and cybersecurity equities and exchange-traded exposure to defense tech, with sentiment skewing toward “resilience and assurance” rather than pure automation. Next, the key indicators are whether militaries publish doctrine updates on human-in-the-loop targeting, auditing of AI outputs, and changes to joint targeting cycle metrics. Watch for policy language that limits AI authority, increases independent verification, or mandates traceability for targeting decisions, because these would directly address the “over-trust” critique. On the cyber side, monitor announcements tied to training pipelines, specialist recruitment, and operational exercises that demonstrate defensive capability under realistic conditions. Trigger points for escalation in this theme include reported incidents where automated systems produced erroneous targeting recommendations, or where cyber operations disrupted data integrity across sensors and networks; de-escalation would look like improved validation frameworks and measurable reductions in false confidence.

Geopolitical Implications

• 01

  Credibility of deterrence and escalation control may weaken if AI-assisted targeting and force assessment lack robust validation.

• 02

  Data integrity and cyber resilience become strategic levers for both defense effectiveness and adversary disruption.

• 03

  Procurement priorities may shift toward assurance, verification, and resilient command-and-control architectures.

Key Signals

• —Doctrine updates on human-in-the-loop targeting authority and AI output auditing
• —Public metrics on joint targeting cycle performance and false-confidence reduction
• —Cyber recruitment/training pipeline announcements and defensive exercise outcomes
• —Any reported incidents linking automation errors to targeting or operational decision failures