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

The Supreme Court's recent ruling on birthright citizenship has sent shockwaves through both legal and political circles. But beneath the constitutional arguments lies a far less discussed story: the technical infrastructure that would need to enforce a policy of "no expectant moms at the border. " As a software engineer who has worked on government data systems, I can tell you that Plan B - the backup strategy reported by Axios - is less about legal theory and more about the messy reality of identity verification, real-time databases. And algorithmic risk scoring.

Behind Trump's reported Plan B lies a hidden architecture of tech systems that would need to identify, track. And ultimately deny entry to pregnant women - a challenge that pushes the limits of modern immigration software.

The concept of "birthright citizenship" has been a pillar of American law for over a century, codified by the 14th Amendment and repeatedly upheld. But when the Supreme Court struck down the administration's executive order on constitutional grounds, the policy machinery didn't stop. Instead, Axios reported that the administration was pivoting to an operational Plan B: using existing border enforcement tools to effectively achieve the same outcome - No expectant moms giving birth on U. S soil.

This is where technology enters the picture. To understand the depth of this shift, we need to examine the data pipelines, biometric systems. And automated decision-making platforms that would execute this policy in practice. Let's break down what "No expectant moms at the border: Trump's birthright Plan B - Axios" actually means from an engineering perspective.

The Legal Backdrop and the Data Challenge

The Supreme Court ruling reaffirmed that anyone born on U. S soil is automatically a citizen, regardless of parental status. Yet the administration's Plan B reportedly focuses on "birth tourism" - the practice of traveling to the U. S specifically to give birth. The legal gray area here is immense, but from a technical standpoint, the challenge is straightforward: how do you build a system that can predict, detect,? And prevent a person's intent to give birth months or weeks in advance?

In production environments, we found that government data systems often struggle with even basic identity matching. The Department of Homeland Security (DHS) relies on the Automated Biometric Identification System (IDENT) - a database that stores fingerprints, facial images. And iris scans from millions of travelers. IDENT is used at ports of entry to verify identities and check against watchlists, but it was never designed to track pregnancy status. Adding a medical variable to an identity database opens a Pandora's box of privacy and accuracy concerns.

Moreover, the sheer volume of data is staggering. According to Customs and Border Protection (CBP) statistics, over 400 million people entered the U. S through ports of entry in fiscal year 2023. Flagging even a fraction of those as potential birth tourists would require algorithms to process travel history, visa types, age, gender. And possibly medical records - none of which are currently integrated into real-time clearance systems.

How Immigration Databases Enforce (or Fail) Birthright Rules

The backbone of U. S immigration enforcement is the Enterprise Service Bus (ESB) that connects USCIS (citizenship and immigration services), ICE (enforcement). And CBP (border) systems. One key component is the Person Centric Query System (PCQS), which attempts to link records across agencies. When a visa applicant or traveler is flagged for "birth tourism," the system would need to pull data from multiple sources: previous visa applications, travel history. And even social media monitoring.

However, as documented in a Government Accountability Office (GAO) report on DHS IT modernization, these legacy systems suffer from fragmentation and inconsistent data quality. In one audit, over 30% of records had missing or conflicting nationality fields. Implementing a rule like "deny entry to visibly pregnant women" would require not only accurate pregnancy detection but also reliable record linkage - a far cry from the current state of affairs.

Furthermore, the concept of "pregnancy" as a data point is fraught with legal and technical ambiguity there's no standard international format for recording pregnancy status in travel documents. Even if CBP officers were given smart glasses or AI-based body scanners to detect pregnancy, the error rates would be high, especially in early stages. False positives could lead to unlawful detentions. While false negatives would undermine the policy's intent.

"No expectant moms at the border" - Automation and Risk Scoring

One plausible technical approach to Plan B is an algorithmic risk-scoring system similar to the one used for visa overstay prediction. Currently, CBP uses a system called the Automated Targeting System (ATS) to assign risk scores to travelers based on factors like travel patterns, country of origin, and prior violations. Extending this to predict "likelihood of giving birth in the U. S. " would require training a model on historical data from known birth tourism cases.

But here's the engineering catch: good training data is scarce. Birth tourism is notoriously hard to quantify because many visitors don't self-report their intent. The most reliable data comes from medical billing records (hospital births paid in cash or foreign insurance), which aren't routinely shared with immigration authorities. Without labeled datasets, any machine learning model would be plagued by bias and low precision.

A 2022 study by the RAND Corporation on AI in immigration enforcement found that risk-prediction models used by DHS had disparate impacts on certain nationalities. A model trained to flag birth tourists might inadvertently target women from specific countries, leading to discrimination lawsuits. The Axios report hints at this tension: the administration wants a policy that effectively achieves "no expectant moms at the border," but the technical tools available are blunt instruments.

Plan B: The Technical Architecture Behind a Shifting Policy

If the legal route is blocked, the fallback is to change operational procedures and software configurations. This is classic "policy as code" - where a rule is implemented through system updates rather than new laws. For example, CBP could update its Primary Inspection Kiosk software to add a mandatory question: "Are you pregnant or planning to give birth in the United States? " The answer would then be stored and used to influence admission decisions.

However, such a change requires significant front-end development, database schema modifications, and backend validation logic. The kiosks run on custom Linux-based systems with Java applications that communicate with the IDENT backend. Adding a new field would require updates across all 300+ ports of entry, rigorous testing to avoid crashes, and training for thousands of officers. Deployment of such a change typically takes 12-18 months for a mature government IT project.

Additionally, the legal department would need to certify that the question doesn't violate medical privacy laws (HIPAA) or anti-discrimination statutes. Civil rights groups would likely challenge the question in court. From a pure software engineering viewpoint, this is a high-risk, low-reward feature that adds complexity to an already brittle system.

Facial Recognition at Ports of Entry - A Case Study

One technology that has been touted as an enabler for Plan B is facial recognition. Currently, CBP uses facial matching at many airports to verify travelers against their passport photos. The system, called Traveler Verification Service (TVS), compares a live camera feed with stored biometric templates. Could it be adapted to detect pregnancy?

Technically, no - not in its current form. Facial recognition algorithms are trained to identify identity, not physiological states. While computer vision models can be trained to detect certain medical conditions (e g., jaundice in newborns), pregnancy isn't easily inferred from facial features alone, especially in early trimesters. A 2023 study by the National Institute of Standards and Technology (NIST) on biased facial recognition showed that even demographic attributes like age and gender are misidentified at double-digit rates for certain populations. Expanding to pregnancy detection would require new labeled datasets and could amplify existing biases.

Moreover, deploying facial recognition for this purpose would raise Fourth Amendment concerns about warrantless searches and First Amendment concerns about chilled speech. The Supreme Court's recent Riley v. California precedent on digital searches suggests that any system capturing biometric data must have robust oversight and clear probable cause standards.

The Data Pipeline for Citizenship Verification

Let's talk about the actual data flow. When a baby is born in the U, and s, hospitals issue a birth certificate that's sent to the state's vital records office. Which then transmits the data to the Social Security Administration and the State Department for passport issuance. Immigration authorities have no direct role in this process. To enforce "no expectant moms at the border," the government would need to create a new data pipeline that connects hospital admissions - specifically, obstetric admissions - with border crossing records.

This would require modifications to the Electronic Verification of Vital Events (EVVE) system, a real-time API used to validate birth records for employment eligibility (Form I-9). Currently, EVVE is read-only and only accessible by employers. Expanding it to allow CBP to query hospital admissions by maternal name would be a massive engineering effort, with implications for data privacy and HIPAA compliance.

From a database architecture perspective, you would need a federated query system that can join CBP travel records with state-level healthcare data. The latency requirement for border clearance is under 30 seconds per traveler. Real-time joins across disparate, privacy-restricted databases at that speed is a hard problem, especially given that many state vital records systems still run on COBOL-based mainframes from the 1970s.

What Software Engineers Can Learn from This Policy Change

For engineers working on government systems - or any system with high-stakes regulatory impact - the birthright policy pivot offers several lessons. First, policy changes often outpace technology readiness. The executive branch can announce Plan B in a press release, but the database schemas, APIs. And UI components needed to enforce it take years to build and validate. As engineers, we must push back against unrealistic timelines and demand realistic feasibility assessments.

Second, bias in data is a fatal flaw, not a bug. Any system designed to predict human intent based on demographic proxies will produce discriminatory outcomes. The Axios story highlights that Plan B focuses on "expectant moms" - a category that's almost entirely female. A policy coded into software without robust fairness constraints will inevitably lead to gender-based screening. Engineers have a responsibility to test for disparate impact using tools like Google's What-If Tool or IBM's AI Fairness 360.

Third, security and privacy aren't afterthoughts. Connecting hospital records to border enforcement creates a single point of failure for sensitive health data. If the pipeline is breached, millions of women's pregnancy histories could be exposed. As we saw with the OPM data breach in 2015, government databases are often prime targets for foreign adversaries. Encrypting data at rest and in transit, applying role-based access controls. And conducting regular penetration tests are non-negotiable.

Ethical Engineering in High-Stakes Government Systems

The discussion around "No expectant moms at the border: Trump's birthright Plan B - Axios" ultimately comes down to ethics in computer science. Every line of code in a border enforcement system has real-world consequences - detentions, family separations. And violations of privacy. The ACM Code of Ethics states that software professionals should "contribute to society and human well-being" and "avoid harm. " Implementing a system to deny entry to pregnant women arguably violates both principles.

Engineers working on such projects must ask hard questions during the design phase: Is this system being built to enforce a law that might be unconstitutional? Is there a less invasive alternative? How can we build in transparency and accountability? These aren't just philosophical questions; they affect whether the system will survive legal challenges and public scrutiny.

Moreover, the concept of "technical impossibility" is often used as a shield by developers. But in this case, the technical challenges - poor data quality, legacy systems, bias, privacy constraints - are real barriers. A responsible engineer would flag these issues in writing, not just let management assume the system can be built on time and on budget.

Frequently Asked Questions

• What is "birthright Plan B" as reported by Axios? It refers to the Trump administration's backup strategy to discourage birth tourism after the Supreme Court struck down executive orders on birthright citizenship. The plan involves using existing border enforcement technologies and procedures to deny entry to women suspected of traveling to give birth in the U. S.
• Can facial recognition really detect pregnancy at the border? No, current facial recognition systems are trained for identity verification, not medical condition detection. They can't reliably detect pregnancy from a face image. And doing so would require entirely new models with high error rates and ethical concerns.
• What databases are involved in enforcing birth tourism rules? Key systems include CBP's IDENT (biometrics), the Automated Targeting System (ATS) for risk scoring, USCIS's Person Centric Query System (PCQS). And potentially hospital vital records databases via the EVVE system.
• How would "no expectant moms at the border" be technically implemented? Likely through software updates to primary inspection kiosks (adding a pregnancy status question), changes to risk-scoring algorithms to flag women of childbearing age, and potentially connecting hospital admission data to border crossing records - all of which pose significant technical and legal hurdles.
• What can software engineers do to ensure ethical implementation of such policies? Engineers should advocate for transparency, conduct algorithmic bias audits, require clear legal authorization for each data point. And document potential harms. They can also refuse to work on features that violate core ethical standards, as outlined in the ACM Code of Ethics.

Conclusion: The Code Behind the Policy

The Axios report on "No expectant moms at the border: Trump's birthright Plan B" reveals a critical truth: modern immigration enforcement is as much about software as it is about law. The Supreme Court can strike down an executive order, but the executive branch can still pivot to operational choices embedded in databases, APIs. And biometric scanners. For those of us who build these systems, the choice is clear - we can either be passive implementers of flawed policies or active participants in designing fair, transparent and technically sound alternatives,

As web developers, data engineers,And software architects, we have a unique vantage point. We see the gaps between policy rhetoric and technical reality. We can calculate the error rates, the latency constraints. And the privacy trade-offs. The question is whether we will speak up before the code is deployed,

Ready to dive deeper Read the full Axios article "No expectant moms at the border: Trump's birthright Plan B" for the political context, then explore the DHS immigration statistics portal to see the actual data behind the debate. If you're an engineer, consider contributing to open-source fairness toolkits like AI Fairness 360 to help build better systems.

What do you think,

Can a government system built on