What Morrow Knew

The public hearing began on a Monday morning in a government building whose heating system clicked loudly enough to remind Elias of the refrigerator. He noticed the association and let it pass.

That, more than anything, told him how much had changed. Morrow's inquiry was not a trial. No one stood in a dock. There were no dramatic objections, no single document that caused the room to gasp. Regulators asked precise questions about research governance, disclosure language, product design and data provenance. Lawyers asked for definitions. Engineers disagreed about versions. Clinicians argued over whether therapeutic benefit could justify reconstructive uncertainty. Most of the work consisted of people explaining decisions that had seemed small when they were made.

Lena testified for nearly four hours. Elias watched from the second row beside Saira Malik. Mara was not present. Her name appeared only as Participant F-17 in documents and as a redacted third party in Elias's Continuity record.

Lena described Fallow's early grief program, the shift from factual fidelity toward Narrative Stability, Mara's objections, and the team's failure to distinguish clearly between a patient feeling coherent and a memory becoming evidentially trustworthy.

"Did you know participants were incorporating generated details into later recollections?" a regulator asked.

"Yes."

"When?"

"We had credible evidence by late 2010."

"And the program continued."

"Yes."

"Why?"

Lena looked down at the documents before her. "Because the same participants often improved clinically. We treated source confusion as a manageable side effect rather than a challenge to the treatment goal."

"Was that reasonable?"

A long pause.

"Parts of it were reasonable at the time. The conclusion we built from those parts was not."

The answer irritated one of Morrow's lawyers, who shifted in his chair and wrote something on a yellow pad. Lena continued before another question came. "I want to be precise. We did not possess technology capable of inserting arbitrary memories into people. We still do not. What we had was more ordinary and, in some ways, more dangerous. We could provide plausible sensory and narrative material during periods when people were actively trying to make sense of incomplete autobiographical events. Repetition increased familiarity. Familiarity influenced confidence. Confidence then became one of the signals later systems used to rank reconstructions."

"A feedback loop."

"Yes."

"Did Morrow understand this before consumer release?"

"Some people did. The company understood pieces of it. That is not the same as saying every person understood the full consequence."

"Did senior leadership?"

Lena looked toward Rohe.

"Enough to require stronger warnings than we gave."

Rohe testified later.

He did not defend Narrative Stability as elegantly as he had in private. Public rooms altered language. He acknowledged approving the removal of persistent uncertainty prompts because users found them intrusive. He acknowledged that source awareness declined afterward. He acknowledged that the early witness-recall pilot had demonstrated increased confidence without comparable gains in accuracy.

"Why was that pilot not disclosed to later users?" the chair asked.

"It was a limited internal study and never commercialized."

"That is not what I asked."

Rohe turned his wedding ring.

Elias almost smiled.

"Because disclosure would have complicated the claim that our primary risk was model error rather than human adoption of model output," Rohe said finally.

The room became very quiet. Not dramatically quiet. Pens still moved. Someone coughed. A ventilation grille rattled.

But the sentence remained.

Morrow's greatest risk had never been that Vela would hallucinate a false scene. It was that a person would live with a plausible scene long enough to stop remembering which parts had been uncertain.

Elias testified on the second day. He refused to enter his reconstructed kitchen into evidence.

The regulator looked surprised. "Mr. Venn, your case is central to the inquiry."

"The relevant technical facts are already documented through system logs and my consent file. The original kitchen recording belongs to my sister. She has not consented to public release."

"Would the recording strengthen your account?"

"Yes."

"And you still decline?"

"Yes."

The lawyer representing Morrow leaned toward his microphone. "Mr. Venn, you are willing to make serious claims about the company while withholding evidence that could contextualize those claims."

Elias looked at him.

A month earlier, the accusation would have frightened him because it threatened completeness. Now he said, "Evidence can be relevant without becoming yours."

Saira did not smile, but he saw her pen stop moving.

The hearing continued.

By evening the first news reports had appeared. Some called Morrow's technology memory manipulation. Others claimed the inquiry proved all reconstructed memories were unreliable. Both statements were wrong in ways that made better headlines than the truth.

Outside the building, cameras waited.

Elias hated being photographed.

He stood at the top of the steps while Tess approached through reporters. "You can use the side exit."

"I know."

"Then why are you here?"

Elias looked at the cameras. For years he had disliked photographs because they felt like evidence captured without permission, little future arguments about what he had been doing and how he had looked while doing it. Now he understood that refusing photographs had never protected him from interpretation.

"I don't want to hide," he said.

Tess nodded.

He walked down the steps. Reporters shouted questions about false memories, Mara, lawsuits, whether Morrow had experimented on him.

Elias answered only one.

"Did they put a memory in your head?"

He stopped.

"No," he said. "That's the wrong question."

Then he kept walking.