Erasure Machines: AI and the Adoption Industry

The Replacement

Artificial intelligence advances with remarkable speed. Companies pour billions into automation, replacing human workers with systems that never tire, never complain, and never ask for a raise.

U.S. venture capital investment in AI infrastructure reached $194 billion in 2025, while enterprise integration rose sharply to 9.2% in Q2 of that year, with select sectors reporting rates of 25% to 30%. The economic logic seems sound. Efficient. Inevitable.

Adoption operates the same way. A child is removed from their family and placed with strangers. The original mother is replaced. The original identity is erased. A new certificate declares a new reality, as if the child sprang into existence on the day of placement. The system calls this progress. Loving. Inevitable.

Both industries traffic in replacement. AI replaces workers with algorithms. Adoption replaces mothers with legal strangers. Both erase what came before. Both profit from the transaction.

Pipeline of Obliteration

When Western demand for adoptable infants exceeded domestic supply, agencies looked abroad. South Korea’s adoption program began with children born to Korean women and Western soldiers. It later expanded to include the children of single mothers and poor families.

The system operated without oversight. Children with living relatives were processed as orphans. Sick infants were swapped with healthier ones to make export quotas. Diplomats from Sweden, Denmark, and Norway pressured Korean officials to maintain the flow of babies, threatening diplomatic consequences if adoptions stopped.

The AI industry operates a similar pipeline. When firms cannot find enough workers or want to cut costs, they look to automation. Analysts report that 80% of business bosses surveyed say autonomous technologies and AI have reduced their workforce, with headcount reductions ranging from 1% to 15%.

The language of “saving orphans” masks a transaction of capital, a transfer of human beings from poor countries to wealthy ones. The language of “efficiency gains” masks the same.

Cost of Speed

Researchers have found that AI-adoption speed itself can determine whether a transition leaves the labour market intact or produces permanent damage.

When AI displaces workers, they enter a retraining pipeline with limited capacity. Faster processing compresses the displacement window: the same mass of workers arrives in much less time, overwhelming the system.

Some workers make a rational calculation to leave the labour force entirely. That exit is absorbing, there is no coming back.

Adoption operates with similar speed. A mother signs away her child. The child is placed with strangers and true identity is hidden away. They grow up with no access to their origins. The speed of the transfer ensures no one asks questions. No one checks on the mother’s coercion or verifies that the child was truly abandoned. The pipeline is too valuable.

Control Problem

Regulators and technologists now debate who controls AI systems. A U.S. bill proposed in 2026 requires advanced AI systems to retain kill-switch capabilities, while an open letter signed by over 1,300 AI workers calls for international cooperation to control the pace of autonomous AI development. The debate centers on a fundamental question: who has the power to stop the machine?

Adoption has no kill switch. No one stops the transfer of a child when fraud is suspected. Anonymous baby boxes remain legal. Most states still seal files forever. Apologies never come, even decades after an adoptee uncovers the truth.

The industry operates without the controls required of any other system that removes human beings from their origins.

Transparency That Should Exist

Adaptive AI systems require continuous monitoring, drift detection, and structured containment when thresholds are breached. Governance experts recommend visibility into system behavior, audit trails, and mechanisms for human override.

These are not radical demands. They are basic safety measures for any system with the power to affect human lives.

Adoption has none of this. No continuous monitoring of real-parent coercion. No drift detection when agencies falsify abandonment papers. No containment when a child is placed with strangers. No audit trails accessible to the person most affected. The adoptee is the only one who cannot see the file.

Reversal Approach

AI advocates warn against too-rapid integration. The Klarna case provides a cautionary tale. The company replaced much of its customer service workforce with an AI system, claiming the assistant handled 2.3 million conversations in its first month alone. Service quality degraded in edge cases. The firm had to reverse course and rehire staff.

Adoption has no such reversal. Once a child is trafficked, their file is sealed. Once the birth certificate is falsified, the original is hidden. Once the honest surname is replaced with a stranger’s, it stays that way. The industry has no need to rehire mothers and never apologizes for its errors.

The AI industry is beginning to understand that speed without oversight creates permanent damage. Human adoption has always understood this. But for them, the damage is a feature, not a bug.