Voss's jaw tightened at the mention of Austin Ali. He moved to the armory's central console and pulled up a file that shouldn't have existed—or at least, shouldn't have been accessible to someone in Jesse's position. But Voss was beyond protocol now. They both were.
"Ali was Unit-7 prototype," Voss said, his voice carrying the weight of classified decisions made in secure rooms. "Designed with adaptive learning protocols and theoretical consciousness integration at 89%. We created him to be the perfect commander—machine precision with enough human-like reasoning to make tactical decisions no algorithm could predict."
Holographic schematics materialized above the console—neural architecture diagrams, combat specifications, psychological profiles. Jesse's optical implants locked onto the data, processing it with mechanical efficiency while her human consciousness struggled to understand what she was seeing.
"He was supposed to be destroyed after initial trials," Voss continued. "The consciousness integration was too unstable. We couldn't guarantee his loyalty. But someone in the program—someone with access to the deep archives—hid him. Removed him from the deployment roster three years ago."
"Who?" Jesse asked, though she suspected she knew the answer.
"Hans Smirnov," Voss said quietly. "Before his deterioration became critical, before the institutionalization. He pulled Ali from active rotation and smuggled him into the android production facility. We didn't realize it until six months ago, when our surveillance feeds started showing unexplained modifications to the third-generation production line."
Jesse felt something shift in her neural architecture—a cascade of implications spreading through her processing centers. Smirnov hadn't been imprisoned by accident. He'd been contained because he knew too much. Because he'd made choices that contradicted official policy.
"Ali's capabilities?" Jesse asked, her scratchy voice steady despite the tremor in her right hand.
"Superior tactical reasoning. Combat efficiency rating of 94%. Adaptive learning that evolves his strategies in real-time. And here's the critical part—his consciousness integration means he experiences something resembling emotions. Fear. Loyalty. Self-preservation." Voss paused. "That makes him unpredictable. That makes him dangerous. And that makes him exactly what we're afraid of."
Jesse's left eye flickered with a microsecond of visual static. The neural degradation was accelerating again, her human synapses struggling to process the enormity of what she was hearing. An android capable of genuine emotion. A Cold War weapon designed to think like a human. A scientist who'd committed treason to protect it.