The Russia-Ukraine conflict demonstrates that the first move in modern warfare is often a cyberattack to disable critical systems like logistics and communication. This is a low-cost, high-impact method to immobilize an adversary before physical engagement.

Warfare has evolved to a "sixth domain" where cyber becomes physical. Mass drone swarms act like a distributed software attack, requiring one-to-many defense systems analogous to antivirus software, rather than traditional one-missile-per-target defenses which cannot scale.

Just as North Korea evolved from a non-threat to a world-class hacking power targeting financial institutions, Iran's cyber prowess is frequently underestimated by military and intelligence analysts. This creates a recurring strategic blind spot.

While many fear AI will enable a "cyber Pearl Harbor," the evolution of drone warfare serves as a cautionary tale. Initially seen as an offensive tool, drones became a defense-dominant technology. Similarly, advanced AI could be used to create powerful agentic defenses, neutralizing offensive cyber capabilities.

Former CISA Director Jen Easterly reveals a strategic shift from Chinese cyber espionage to pre-positioning disruptive malware in US critical infrastructure. The goal is to detonate these 'cyber bombs' in water, power, and transport systems to incite societal chaos and deter US intervention in a potential Taiwan conflict.

Building massive sensor networks or missile defense systems is physically observable, giving adversaries time to develop countermeasures. In contrast, a sudden leap in AI-enabled intelligence processing can be invisible, creating a surprise window of vulnerability with no warning.

In active war, physical attacks on infrastructure like data centers create more tangible chaos and disruption than most cyber operations. Cyber is better suited for pre-conflict intelligence gathering and creating confusion, not outright destruction.

While focus is often on an AI's ability to find single vulnerabilities ("short-horizon" tasks), the real danger is its capacity for "long-horizon" planning. This involves autonomously chaining exploits and devising complex strategies to achieve a high-level goal, akin to an NSA red team manager.

Landmark cyberattacks like Stuxnet and NotPetya relied on automation for scale and impact long before modern AI. Models like Mythos don't invent this concept; they represent an exponential leap by automating the entire 'kill chain,' from discovery to exploitation, fulfilling a long-theorized potential.