No expectant moms at the border: Trump's birthright Plan B - Axios

The news broke with a headline that felt more like policy science fiction than real governance: No expectant moms at the border: Trump's birthright Plan B - Axios. The Supreme Court had just overturned the administration's attempt to unilaterally redefine birthright citizenship via executive order. And the response from the White House was swift-a "Plan B" that targets birth tourism, visa overstays. And the very software systems that record who is a citizen. This policy shift isn't just a legal battle-it's a stress test for the nation's entire civic identity infrastructure. As a software engineer who has worked on identity verification systems for government agencies, I see this as a profound moment where law, ethics. And engineering collide.

To understand the stakes, you have to look past the political rhetoric and into the database schemas. The Fourteenth Amendment guarantees citizenship to anyone born on U. S soil. But enforcing that guarantee means having a reliable, auditable digital record of every birth. The Trump administration's Plan B-as reported by Axios-doesn't try to amend the Constitution; instead, it aims to make birthright citizenship functionally unattainable by weaponizing the very technology that processes birth certificates, passports. And visa applications. This is where the story gets deeply technical.

From Executive Order to Database Query: The Real Battlefield

The Supreme Court ruling that struck down the initial birthright citizenship restrictions was a legal milestone. But it didn't touch the software. The administration's Plan B, according to Axios and corroborated by The Guardian, focuses on "cracking down on birth tourism"-prosecuting companies that market U. S delivery packages and denying visas to pregnant women. From a systems engineering perspective, this requires modifying the consular visa processing algorithms, updating the Customs and Border Protection (CBP) lookout databases. And re-training the AI models that flag suspicious travel patterns.

In production environments, we've seen how quickly well-intentioned policy changes can break legacy systems. The Department of Homeland Security's HART (Homeland Advanced Recognition Technology) system. Which manages biometric identification, was already strained. Adding a filter for "pregnant traveler" or "likely to give birth in the U. S. " would require not just new business rules but entirely new data pipelines. The technical challenge is immense: how do you define "expectant mom" in a way that a machine can reliably classify without violating privacy or medical ethics? You can't. Every heuristic bleeds into false positives and racial profiling.

The Data Engineering of Birthright Citizenship: A Case Study in Integrity

Birthright citizenship is, at its core, a data integrity problem. Every U, and s-born citizen gets a Social Security number (SSN) tied to a birth certificate. The National Center for Health Statistics reports over 3, and 6 million births annuallyEach record must link to a parent's citizenship status-and that status is fluid. A lawful permanent resident today might become a citizen tomorrow, retroactively changing the child's legal classification. Current systems weren't designed to handle that dynamism.

During my time working on a state-level vital records modernization project, we found that birth certificate databases rarely store the parent's immigration status in a machine-readable field. Instead, it's buried in PDF attachments, paper forms, or textual notes. To enforce a policy like Trump's Plan B, you'd need to extract, normalize, and cross-reference that data against CBP arrival records, USCIS case status files. And even credit bureau databases. That kind of ETL (extract, transform, load) pipeline is prone to errors, latency. And privacy leaks. The administrative burden alone could overwhelm the SAVE program (Systematic Alien Verification for Entitlements),, and which handles citizenship verification for public benefits

Biometric Black Holes: How Facial Recognition Could Enforce Plan B

One of the creepier aspects of the data mining Plan B is the use of biometrics at borders. CBP already uses facial recognition cameras at ports of entry. The technology matches travelers against a gallery of visa-holders and watchlists. Administration officials have suggested that these systems could be used to track pregnant women entering the U. S and flag them for secondary inspection. The problem is that pregnancy isn't a biometric trait-there's no algorithm that can reliably detect due date from a photo. Even if you could, would you want the government storing medical data collected without consent?

In a 2023 audit, the Government Accountability Office found that CBP's facial recognition systems had a 96% match rate for U. S citizens but only around 85% for non-citizens, with significant disparities across skin tones and age groups. Deploying such a system for selective enforcement against an invisible condition like early pregnancy would be a recipe for civil rights violations. As engineers, we know that when you try to solve a social problem with technology, you often just automate the bias.

The Software Stack Behind Citizenship Verification (It's Scarier Than You Think)

Most people assume that verifying citizenship is a simple database lookup. In reality, the stack is a patchwork of COBOL-era mainframes, Java APIs from the 2000s, and modern cloud microservices that don't talk to each other. The US-VISIT program. Which tracks entry and exit, runs on a system originally built in 2004. The Electronic Verification of DOJ-Related Documents (E-Verify) API has endpoints that return ambiguous results like "Temporary Non-Confirmation" with no explanation. Now imagine adding a rule that says "If mother wasn't a citizen at the time of birth, flag child as non-citizen for passport application. " The E-Verify system doesn't even store the mother's status; it only checks work eligibility.

To implement Plan B, the government would need to create a new data type: "birthright citizenship status derived from parent. " That would require changes to the SSA's Numident database, the State Department's Consular Consolidated Database, and USCIS's Central Index System (CIS). These are mission-critical systems with uptime requirements of 99. 99%. Rolling out a schema change without breaking downstream passport printing, driver's license issuance, and Social Security benefits would take years and cost billions. The Axios reporting suggests the administration is looking for executive action shortcuts-but shortcuts in database migrations are exactly how we get data corruption and wrongful denials.

Blockchain and the Immutable Ledger Fantasy for Citizenship

Whenever a data integrity problem arises, someone inevitably suggests blockchain. The idea: store birth records on an immutable distributed ledger, so that no future administration can retroactively change citizenship status. Estonia has done something similar with its e-Residency program. But for permanent legal identity, not birthright. The challenge is that citizenship isn't a single atomic event-it's a relationship between a person and a state that can change over time. A blockchain recording birth citizenship can't capture the fact that a parent naturalizes later, potentially altering the child's derivative status. Furthermore, blockchain is only as trustworthy as the oracle that submits the data. If a border agent misclassifies a parent, the error is immortalized.

I've seen multiple proposals for "citizenship on the blockchain" in academic papers, but none address the governance question: who controls the keys? The U. S government would never delegate control of such a critical register to a decentralized network. And if it's a permissioned blockchain, you've just added overhead without solving the core problem of accurate data entry. The real solution is better data validation at the point of entry-ensuring that hospital birth clerks can easily query immigration status via a secure, real-time API that's a software engineering challenge, not a cryptographic one.

Ethical Failure Modes in Automated Citizenship Decisions

Let's zoom out. The Trump administration's Plan B is a case study in how not to build civic tech. When you prioritize speed over robustness, you inevitably produce false negatives-people who are actually citizens but are denied passports or benefits because a system says otherwise. During the 2020 census, we saw how data glitches could undercount minority populations and cost states congressional representation. A citizenship database error is worse: it strips someone of their fundamental rights with no easy recourse.

As engineers, we have an ethical responsibility to question the use case. The Association for Computing Machinery (ACM) Code of Ethics says that "computing professionals should contribute to the development of systems that support the public good. " Implementing an algorithmic system designed to disenfranchise a specific group-pregnant women at the border-fails that test. Even if the code is legally permissible, it's technically flawed and morally dubious. The Supreme Court's ruling didn't just strike down the order; it implicitly said that some rights are too fundamental to be delegated to an algorithm.

What Developers Should Know About the Immigration Tech Stack

If you're a software engineer reading this and thinking about government contracting, here are the practical takeaways. First, any system that touches citizenship classification needs an audit trail with granular detail: who made the determination, what data they used. And how they overrode defaults. The OpenAPI specification for USCIS's APIs is publicly available. And it's instructive to see how few endpoints handle "citizenship derived from parent" as a first-class field. Second, never hardcode policy logic. Use feature flags and externalized rules engines so that courts can review policies without rewriting software. Third, build in a human-in-the-loop for any decision that could revoke or deny a benefit. The E-Verify system already allows for "tentative non-confirmations" that require manual follow-up; that safety valve should be standard across all citizenship verification flows.

Finally, consider the data retention implications. If a birth certificate is flagged because the mother was on a temporary visa, that flag should have an expiration date tied to the mother's eventual naturalization. Without clean-up jobs, you'll get permanent stigma attached to records. I've seen Medicare systems that still had "dead" flags from system glitches in the 90s. Citizenship data must be treated as living, not static.

Frequently Asked Questions

1. What exactly is "birthright citizenship" and how does it work technically? Birthright citizenship is governed by the Fourteenth Amendment: anyone born on U. S soil is a citizen. Technically, it's recorded on a birth certificate,, and which then triggers an SSN and passportThe system relies on the mother's self-reported citizenship at the hospital. Which is rarely verified in real-time.
2. How would Plan B change the engineering of citizenship records? It would require adding a new field to birth records indicating the mother's immigration status at the time of birth. Which then must be cross-checked against CBP entry/exit data and USCIS case files-a massive data integration effort.
3. Can AI accurately detect pregnancy from border video feeds? No, and current facial recognition can't measure gestational ageAny claim of detecting "expectant moms" is based on travel patterns or medical records, not video inference. And raises serious privacy and consent issues.
4. What are the biggest technical risks of this policy? False negatives (citizens denied passports), data corruption from schema changes in legacy systems. And unilateral algorithmic bias against foreign-born parents and their children.
5. Is there a precedent for blockchain-based citizenship records? Estonia uses blockchain-backed identity for e-residency, but not for birthright citizenship. No country has a fully decentralized, immutable ledger for citizens, primarily due to governance and data accuracy challenges.

Conclusion: The Real Plan B Is Better Engineering

The story of "No expectant moms at the border: Trump's birthright Plan B - Axios" isn't just a political saga; it's a cautionary tale for anyone building software that touches civil rights. The impulse to solve policy problems with technology is understandable, but when the technology is brittle, the data is dirty, and the ethics are ignored, you end up with systems that cause real harm. Instead of patching workarounds, we need to invest in modernizing the underlying identity infrastructure-secure APIs, real-time verification, clear metadata standards. And human-in-the-loop oversight.

If you're a developer or architect working on civic systems, I challenge you to read the Axios piece on Plan B, then review the public documentation for USCIS's integration points. Raise questions about data lineage, error rates, and failure modes before the next administration tries to flip the switch. The code we write today determines whose rights are recognized tomorrow.

What do you think?

Should immigration enforcement software be subject to mandatory algorithmic impact assessments before deployment, similar to the AI Bill of Rights framework?

If you were the CTO of DHS, how would you architect a system that allows birthright citizenship verification without creating a permanent data shadow for children of non-citizen parents?

Is it ethical for an engineer to knowingly build a system whose sole purpose is to reduce the number of people eligible for birthright citizenship-even if the law allows it?