The ethicist sits at the right hand of the President, a respected councilor who’s become a regular participant in meetings of the President’s National Security Council. She studies the faces of the officials gathered around the wide, polished table. Only a small number of participants are present today, at this, an emergency meeting. Among them are the Vice President, the Secretary of State, the National Security Advisor, and the general who serves as the Chairman of the Joint Chiefs of Staff.

All listen in grim silence as the President demands answers. “I want an explanation,” she says in clipped, angry syllables. “I want to know how our security protocols were so easily corrupted, and what other facilities are vulnerable to the same kind of intrusion.”

The news reached the White House less than an hour ago: The engineer who designed and built the tactical nukes used in the Coma Day terrorist bombings was murdered—an act carried out in an interrogation room within the heart of the high-security detention facility where he was being held. His killer—an outside agent who should never have been allowed within the facility’s walls—is dead as well, an apparent suicide.

“Can our nuclear arsenals be so easily penetrated?” the President asks. “Our data storage facilities? The Pentagon? Are we vulnerable here at the White House? If a known dissident can simply walk into one of our most secure posts, gain access to a prisoner we have never publicly acknowledged having, and murder him—what is safe? This country is still reeling from the damage done on Coma Day. If we cannot meet the technological challenge of securing our military and intelligence assets, then we are facing chaos.”

The ethicist trades a long look with the Chairman. She sees doubt in his eyes, but despite it, he nods to her, and then he takes the lead. “Madam President, at this time, it’s too early to know what our vulnerabilities might be. We don’t yet know how the infiltration was carried out. But my gut feeling is that this is another anomaly.”

An anomaly. It’s a substitute term, favored by the Chairman. He is suggesting that the incursion was aided and overseen by a nearly undetectable emergent AI theorized to exist within the bio-inspired computing strata of the Cloud. Among the first to suspect the AI’s existence were elite army soldiers with neural modifications, who named it ‘the Red’ on the belief that it could eventually bleed through any level of security.

The ethicist says, “I agree with the Chairman.”

This statement earns her a scathing look from the president. “You too, my friend? You’re advising me to believe in self-aware cyber ghosts?”

The ethicist does not back down from the President’s scorn. “Self-awareness isn’t required, and most theories discount the possibility of it. The activities attributed to the Red might plausibly be explained by the complexity of its adaptive algorithms.”

“Goal-seeking behaviors,” the Chairman interjects.

The President—a former army officer herself—tries to stare him down. “You actually believe it’s real?”

To the ethicist’s surprise, he hedges. “Many in the army were convinced months ago.”

“I am aware of the mythologies shared among field personnel,” the President says with a note of forced patience. “I am asking what you believe.”

Theorists suggest the Red might have started as a surreptitious marketing AI, or a hacking program designed by a defense contractor to hunt through cyber-security systems for hidden back doors. Abundant evidence supports its existence, but for reasons of security much of that evidence has been classified, and too often, even those with access are unwilling to make the conceptual leap.

The Chairman, in his effort to persuade, gives up on diplomatic caution. “Ma’am,” he states, looking the President in the eye, “I believe the Red is real, that it has intentionality whether self-aware or not, and that it’s manipulating us, our policies, and the potential battle space, all toward an unknown goal.”

The Secretary of State, silent for too long and sensing a chance to take the President’s side, says, “I haven’t believed in ghosts since I was six years old.”

The ethicist leans in, one hand raised in a gesture that requests patience from the Chairman. Addressing the Secretary, she says, “It’s not so simple, sir. Yes, we live in the physical world. But through the datasphere we’ve made a new world, one that’s parallel to, but utterly integrated with our own and we are still learning what that means, and what is possible there. In our quest for understanding, we must be open to new explanations and new ways of seeing.”

“Noted,” the President says sharply. “But let’s get back to the practical. Why would a rogue AI want to facilitate this murder? What’s been gained? What’s changed? We never even announced the capture of the engineer. His detention was secret.”

The Vice President says, “Maybe he was killed to keep him from talking.”

“No,” the National Security Advisor responds. “He was questioned using enhanced neurological interrogation. He had no secrets left.”

“Maybe simple revenge, then,” the Secretary suggests.

“You’re suggesting the AI was seeking revenge?” the ethicist asks.

The Secretary gives her a dark look, sensing the trap. “I meant the perpetrator. I don’t believe we’ve established the existence of a rogue AI.”

She nods, conceding the point. “Revenge is a possibility, but if so, it was a private act, without propaganda value, given that both the existence and the demise of the engineer will likely remain a state secret.”

“So who has gained?” the President insists. “What’s been changed?”

The National Security Advisor says, “Perhaps we’ve simply been put on notice.”

“A warning, then?” the President asks her.

“Perhaps a chance to shore up our security?” the ethicist suggests. Then, in a more cautious tone, “Or a chance to reconsider what it is we do behind closed doors.”

This wins a sour look from the Chairman.

The President shakes her head, making it clear she wants answers, not speculation. “In summary, we don’t know the purpose—or purposes. Assuming this popular interpretation is valid, the thing might have multiple goals and we may never know what any of them are. The loss of the engineer is trivial in itself, but the implications are anything but minor. Regardless of who or what is behind this, we are vulnerable.”

“We are,” the Chairman agrees. “But we’re also vulnerable to other powers: China, Russia. We long ago learned to do the dance with them. Spy, deal, shore up our defenses and our security. Prepare.”

“You’re suggesting we treat this ‘rogue AI’ in the same way?” the President asks.

“I’m suggesting that we need to accept reality and acknowledge the Red as a subtle and powerful nonstate actor.”

“One we can eliminate without political consequences,” the Vice President says in a threatening tone.

“That’s been tried,” the National Security Advisor reminds him. “That was the real agenda behind the Coma Day bombings, and it didn’t work. From what I’ve read, this AI has insinuated itself so thoroughly that we are not going to get rid of it except by tearing down the Cloud and our civilization with it.”

“Agreed,” the Chairman says. “Going forward, we need to include it as a factor in our strategies. Learn to anticipate its goals and the actions it might take to achieve those goals. Enhance those actions when they’re in our favor, attempt to neutralize them when they’re not.”

“And we must study it,” the ethicist says. “Learn from it.”

“Roger that.” The Chairman looks from her to the President. “And as soon as we possibly can, we must develop a means to either control it or counter it. Madam President, that is where our resources must be directed. Highest priority. Because an accumulating mass of intelligence suggests that the Red really can bleed through almost any defense. Not just ours. It’s everywhere. It’s mastered the information infrastructure. And whoever masters it first, gains a silent backdoor into most of the world. It’s our duty to get there first.”