Chapter 155: A New Kind of Theft

Chapter 155: Chapter 155: A New Kind of Theft

The email that reopened everything arrived exactly one week after Haruki’s speech at his old school gate, and the timing felt less like coincidence and more like the universe’s particular, unfunny sense of humor about how long peace was actually allowed to last.

*Subject: Urgent — Unauthorized AI Training on Creator Works. Need Lantern Project’s Help.*

He almost deleted it, the subject line carrying the particular breathless urgency that usually meant either genuine crisis or well-intentioned overreaction. Something about the sender’s name, though — a mid-tier creator he vaguely recognized from industry panels — made him open it properly instead of dismissing it alongside the usual daily flood of similar requests.

The email’s contents sobered him considerably faster than he expected.

*A major AI development company has been scraping and training generative models directly on creator portfolios without consent, compensation, or even notification. I discovered my own work embedded in outputs from their public demo three days ago. When I contacted them, their legal team cited a licensing clause buried in a platform terms-of-service update from eighteen months ago that apparently granted them exactly this permission, without ever clearly explaining what we were actually agreeing to.*

Haruki read the email twice, then a third time, the familiar cold recognition settling into his chest — not the exact same shape as Horizon’s retention clauses or Ariadne’s predatory contracts, but unmistakably built from the same underlying philosophy. Find language complicated enough that creators wouldn’t fully understand what they were signing away. Bury the actual mechanism deep enough that by the time anyone noticed, the damage had already been thoroughly, quietly done.

He called Suzune immediately, who answered with the particular weary alertness of someone who’d learned, over years of exactly this kind of call, never to assume a quiet week would last.

"Tell me you’re calling about something boring," she said.

"I wish. Someone found their own creative work embedded in an AI model’s training data. Buried consent clause, similar structure to what we’ve seen before, except this time it’s an entirely different kind of exploitation."

Silence, then a long exhale. "How widespread?"

"I don’t know yet. But if one platform did this quietly, there’s no reason to assume it’s isolated."

By that afternoon, the Lantern Project’s core team had reconvened for the first time in months, the familiar conference room at Yamashita Publishing filling once more with faces that carried the particular weathered readiness of people who’d fought this exact shape of battle before, just never quite this specific version of it.

"This is genuinely new territory," Nakamura said, reviewing the initial evidence spread across the shared screen. "Predatory contracts, we understand intimately. This is something different — creative work being consumed and repurposed at scale, without the kind of clear individual contract we’re used to identifying and challenging."

"The mechanism’s different," Rika said, her expression sharp with the particular focus she brought to genuinely complex problems. "But the underlying exploitation feels familiar. Powerful companies burying consent in language deliberately designed to obscure what creators are actually giving up."

Hayato, who’d joined the meeting remotely from Kanzaki Holdings’ office, leaned forward on the video call. "I can have my legal team start reviewing similar terms-of-service agreements across major platforms immediately. If this is happening quietly at one company, I’d want confirmation whether it’s an industry-wide pattern before we decide how aggressively to respond."

"Agreed," Haruki said. "But we need to move carefully. This isn’t the same kind of villain we’ve fought before. AI development companies have genuine institutional power now, considerable public and political support in some circles, and legal frameworks that haven’t caught up to the actual harm being done. We can’t just apply the Horizon playbook directly."

The meeting stretched for hours, the coalition working through the particular unfamiliar complexity of a challenge that resembled their previous fights in spirit while requiring an entirely different tactical approach. By evening, a preliminary plan had emerged — technical investigation into the scope of the problem, careful legal analysis of exactly what consent had actually been obtained versus buried in incomprehensible terms, and outreach to affected creators who might not even realize their work had been used this way.

That night, Haruki called Yui, walking her through the day’s developments while Hana’s quiet breathing carried faintly through the phone from wherever she’d finally settled for sleep.

"I thought we were past this," Yui said, though her tone carried understanding rather than complaint. "I thought the last fight was supposed to be the last one."

"I think I’ve finally learned there isn’t actually a last one," Haruki admitted. "Not permanently. New exploitation finds new forms. New technology creates new opportunities for the same underlying pattern — powerful people taking advantage of people who don’t have the resources to protect themselves properly."

"Does that discourage you?"

He considered the question seriously, the way he’d learned to over years of practice. "No, actually. It used to, I think. Now it just feels like the actual, ongoing shape of the work. Not a single battle to win and then rest forever. A continuous choice to keep showing up for people who need protecting, however the specific threat happens to look this time."

"That’s a healthier way to think about it than you would have managed five years ago," Yui said warmly.

"I had good teachers," he said, the familiar phrase settling comfortably into its established place in their ongoing conversation.

"You keep saying that."

"Because it keeps being true."

They talked for a while longer, the conversation eventually drifting toward ordinary things — Hana’s latest drawing attempts, Yui’s own expanding outreach work, the small, comfortable details of a life built steadily over years of deliberate, honest effort. Eventually, after they said goodnight, Haruki sat alone in his study, staring at the initial evidence Suzune had compiled about the AI training issue, feeling the familiar, complicated mixture of dread and purpose that always accompanied the beginning of a genuinely difficult fight.

His phone buzzed one final time before he finally settled into sleep. Sora, characteristically blunt even at this late hour.

*Heard about the AI thing. Count me in. Also, did you know Rika already drafted three different legal strategies before dinner? She’s terrifying. I love her. Anyway. We’re doing this, right?*

Haruki smiled, typing back a quick confirmation, then set the phone aside and looked out his study window at Tokyo’s endless, familiar night skyline.

Ten years since that first system window had unfolded in front of him outside a school gate. A career built, a family raised, an entire industry meaningfully reformed through years of stubborn, collective, honest effort.

And now, apparently, an entirely new frontier of exploitation waiting to be understood, challenged, and hopefully, eventually, reformed the same careful way everything else had been.

The fight wasn’t finished. Perhaps it never actually would be, not completely.

But for the first time in a long while, Haruki felt something steadier than dread settling into place alongside the familiar weight of a new challenge.

He knew, now, exactly how to build what this particular fight would require.

He simply had to begin.