AI Ethics
•
8/28/2025
•
19 min read
•
2854 views

AI, Migration, and Border Justice: Technology, Surveillance, and the Fight for Human Mobility Rights

EU border tech budget €113.3B. AI drones in 15+ countries. 3,000+ deaths annually. Comprehensive guide to border surveillance, EU AI Act gaps, legal challenges, and advocacy.

C

By Compens AI Research Team

Insurance Claims Expert

AI, Migration, and Border Justice 2025: Technology, Surveillance, and the Fight for Human Mobility Rights

Comprehensive Guide Updated: December 2025

Quick Overview

| Key Information | Status 2025 | |-----------------|-------------| | EU Border Tech Budget | €113.3 billion (2021-2027) | | Budget Increase | +94% vs previous cycle | | AI Drones Deployed | Active in 15+ EU countries | | Biometric Systems | 12 EU states testing | | Annual Border Deaths | 3,000+ documented (Mediterranean) | | ETIAS Launch | Mid-2025 |

---

Table of Contents

---

The Expanding AI Border Complex {#border-complex}

AI is transforming global immigration enforcement and border control in unprecedented ways. From the EU's Frontex operations to the US-Mexico border, technology is increasingly used to surveil, track, and restrict human mobility—often with significant human rights implications.

The Global Border Industrial Complex

The border security industry has grown into a multi-billion dollar global market:

Market Size and Growth:
  • •Global border security market: $65+ billion annually
  • •Projected growth: 7.5% CAGR through 2030
  • •AI/ML segment: Fastest growing at 15% annually
  • •Key players: Airbus, Thales, Leonardo, Elbit Systems, Palantir
Key Trends:
  • •Increased automation of border decisions
  • •Expansion of biometric databases
  • •Integration of AI with existing surveillance
  • •Growing use of predictive analytics
  • •Privatization of border enforcement

Why This Matters

The expansion of AI at borders affects:

  • •Asylum seekers: Algorithmic screening of protection claims
  • •Refugees: Surveillance in camps and transit
  • •Migrants: Predictive policing and enforcement
  • •Citizens: Expansion of surveillance infrastructure
  • •Communities: Chilling effects on immigrant neighborhoods
  • •Democracy: Normalization of mass surveillance

---

EU AI Border Surveillance Technologies {#eu-surveillance}

Current Deployment Status

According to Euronews and other sources, the EU is deploying a vast array of AI-driven technologies at its borders:

| Technology | Status 2025 | Countries | |------------|-------------|-----------| | AI-powered drones | Active deployment | Greece, Spain, Italy, Poland | | Thermal imaging cameras | Widespread | All external border states | | Biometric systems | Testing/Active | 12+ EU states | | Automated Border Control (ABC) | Active | Major airports EU-wide | | ETIAS screening system | Launching mid-2025 | All Schengen states | | High Altitude Pseudo Satellites (HAPS) | Testing | Mediterranean region | | Underwater sensors | Active | Maritime borders | | AI-powered CCTV | Expanding | Border crossing points |

Budget and Funding

The EU's 2021-2027 framework increased border policy funding by 94%:

| Fund | Amount | Purpose | |------|--------|---------| | Border Management | €6.24 billion | Technology and operations | | Frontex | €5.6 billion | Agency operations | | eu-LISA | €2.1 billion | IT systems | | Horizon Europe | €500+ million | Border research | | Total | €113.3 billion | Security and borders |

ETIAS: The New Screening System

The European Travel Information and Authorisation System launching in 2025:

What ETIAS Does:
  • •Pre-screens visa-exempt travelers before arrival
  • •Uses AI to assess security and migration risks
  • •Cross-references multiple databases automatically
  • •Issues travel authorization or flags for review
Concerns:
  • •Algorithmic profiling based on nationality
  • •Limited transparency in decision criteria
  • •No meaningful appeal process
  • •Potential for discriminatory outcomes

---

Frontex Technology Programs {#frontex}

Agency Overview

Frontex (European Border and Coast Guard Agency) has become one of the EU's largest and most controversial agencies:

Key Statistics:
  • •Staff: 10,000+ (projected)
  • •Budget: €845 million (2024)
  • •Operations: 15+ countries
  • •Standing Corps: Building toward 10,000

2025 Technology Initiatives

May 2025: Long-endurance tactical drone pilot with Bulgaria
  • •Partnership with defense contractors
  • •24+ hour surveillance capability
  • •Maritime and land border coverage
June 2025: Live demonstrations in Burgas, Bulgaria
  • •AI-powered detection systems
  • •Integration with ground forces
  • •Real-time data processing
September 2025: Finland prototype showcase including:
  • •Advanced data fusion systems
  • •Digital twins of border regions
  • •Next-generation sensors
  • •Autonomous maritime vessels
  • •Predictive analytics platforms

Frontex Surveillance Assets

| Asset Type | Number | Capability | |------------|--------|------------| | Aircraft | 20+ | Aerial surveillance | | Drones | 50+ | Long-range patrol | | Vessels | 30+ | Maritime operations | | Vehicles | 100+ | Land operations | | Fixed sensors | 1000+ | Continuous monitoring |

Accountability Concerns

Frontex has faced significant criticism:

  • •Pushback allegations: Documented involvement in illegal returns
  • •Fundamental rights violations: Inadequate oversight mechanisms
  • •Lack of transparency: Secretive operations and contracts
  • •Democratic deficit: Limited parliamentary control
  • •Privatization: Extensive contractor relationships

---

The EU AI Act's Migration Gaps {#ai-act}

What the AI Act Allows

Despite the AI Act's entry into force in 2025, critics argue it fails to protect people on the move:

| Provision | Impact on Migrants | |-----------|-------------------| | Border zone exemption | "Publicly accessible spaces should not include prisons or border control" | | High-risk classification | Allows risk assessment without prohibition | | Automated processing | Permits algorithmic decision-making in asylum | | National security exception | Broad exemptions for border enforcement | | Law enforcement carve-outs | Reduced protections at borders |

Specific Failures

1. Biometric Surveillance Exception

The AI Act bans real-time remote biometric identification in public spaces BUT exempts:
  • •Border control zones
  • •Immigration enforcement
  • •National security operations
  • •"Serious crime" prevention

2. High-Risk Without Prohibition

Border AI systems are classified as "high-risk" but NOT prohibited:
  • •Risk assessments for asylum claims
  • •Credibility assessments
  • •Lie detection technologies
  • •Emotion recognition systems

3. Limited Individual Rights

Migrants have fewer protections than EU citizens:
  • •No right to explanation for AI decisions
  • •Limited access to data held about them
  • •Restricted legal standing to challenge decisions
  • •No effective remedy mechanisms

Documented Concerns from Legal Experts

According to Verfassungsblog and other legal analyses:

  • •Algorithmic bias: Systems exhibit racial and national origin discrimination
  • •Lack of oversight: Inadequate supervision leads to unfair decisions
  • •Vulnerable populations: Migrants often in particularly vulnerable positions
  • •Due process: Limited human review of automated decisions
  • •Accountability gap: Unclear liability for AI-caused harm
  • •Data protection: Inadequate safeguards for sensitive information

---

US Border Technology {#us-border}

Current US Border AI Systems

| System | Function | Status | |--------|----------|--------| | Autonomous Surveillance Towers | AI detection | 200+ deployed | | Integrated Fixed Towers (IFT) | Radar/camera arrays | Active in AZ | | RVSS (Remote Video Surveillance) | AI video analysis | Expanding | | CBP One App | Asylum scheduling | Mandatory for many | | Biometric Entry/Exit | Facial recognition | All major ports | | Predictive Analytics | Migration forecasting | Active |

CBP Contracts and Spending

Major Technology Contractors:
  • •Anduril Industries: Autonomous towers
  • •Palantir: Data analytics
  • •General Atomics: Drones
  • •Elbit Systems: Surveillance
  • •General Dynamics: IT infrastructure

Annual Technology Spending: $2.5+ billion on border technology

Documented Problems

Autonomous Tower Issues:
  • •High false positive rates in harsh weather
  • •Indigenous land violations
  • •Environmental destruction
  • •Militarization of communities
  • •Cost overruns
CBP One App Problems:
  • •Discriminatory facial recognition (fails on darker skin)
  • •Crashes and glitches preventing asylum access
  • •Digital divide excluding vulnerable populations
  • •Geolocation tracking concerns

---

Technologies Currently Deployed Worldwide {#technologies}

Surveillance Systems

| System | Function | Locations | |--------|----------|-----------| | AI drones | Aerial tracking | EU, US, Australia, Israel | | Thermal cameras | Heat detection | Most land borders | | Motion sensors | Movement alerts | Remote areas worldwide | | Facial recognition | Identity verification | Airports, crossings globally | | Predictive analytics | Route prediction | Multiple regions | | Satellite imagery | Area monitoring | Maritime, desert regions | | Underwater sensors | Maritime detection | Mediterranean, English Channel | | Ground-penetrating radar | Tunnel detection | US-Mexico border |

Database Systems

EU Systems:
  • •Eurodac: Fingerprint database for asylum seekers (4.9 million records)
  • •SIS II: Schengen Information System (90+ million alerts)
  • •VIS: Visa Information System (70+ million applications)
  • •ETIAS: European Travel Information and Authorisation System (new)
  • •EES: Entry/Exit System (launching)
  • •ECRIS-TCN: Criminal records for third-country nationals
US Systems:
  • •IDENT/HART: Biometric database (270+ million identities)
  • •TECS: Law enforcement database
  • •ATS: Automated Targeting System
  • •NSEERS: (Suspended but data retained)

Emerging Technologies

Currently in Development or Testing:

  • •Lie detection AI: Analyzing micro-expressions, voice patterns
  • •Emotion recognition: Assessing "deception" or "stress"
  • •Gait analysis: Identifying individuals by walking pattern
  • •Social media analysis: Screening based on online activity
  • •Predictive policing: Forecasting migration flows
  • •Autonomous vehicles: Self-driving patrol vehicles
  • •AI translation: Real-time interview translation with sentiment analysis

---

Human Rights Implications {#human-rights}

Documented Impacts

1. Deaths at Borders

Technology pushes migrants to more dangerous routes:

| Region | Deaths 2024 | Trend | |--------|-------------|-------| | Mediterranean | 3,000+ | Increasing | | US-Mexico border | 800+ | Increasing | | English Channel | 50+ | Increasing | | Sahara Desert | 500+ (estimate) | Unknown |

2. Forced Dangerous Routes

Surveillance creates deadly detours:
  • •Desert crossings instead of monitored paths
  • •Dangerous sea routes to avoid detection
  • •Longer journeys through hostile terrain
  • •Reliance on smugglers for technology evasion

3. Pushbacks and Refoulement

AI systems facilitate illegal returns:
  • •Rapid processing enables pushbacks before claims heard
  • •Automated systems override individual assessment
  • •Technology creates distance from accountability
  • •Data sharing enables return to danger

4. Family Separation

Automated enforcement impacts families:
  • •Algorithmic decisions without humanitarian consideration
  • •Biometric systems track and separate family members
  • •Children detained based on automated flags
  • •Parents deported while children processed separately

5. Community Surveillance

Internal enforcement extends border:
  • •License plate readers in immigrant neighborhoods
  • •Facial recognition at community locations
  • •Data sharing between agencies
  • •Chilling effect on accessing services

6. Due Process Violations

Algorithmic decisions without appeal:
  • •No explanation of AI-based denials
  • •Limited human review of automated decisions
  • •Language barriers in challenging outcomes
  • •Inadequate legal representation

Academic Research Findings

A Cambridge study found that AI border technologies create a "subtle erosion of human rights" through:

  • •Opacity: Decisions made by black-box algorithms
  • •Accountability gaps: Unclear responsibility for harm
  • •Disproportionate impact: Vulnerable groups most affected
  • •Limited recourse: Few mechanisms to challenge AI decisions
  • •Normalization: Surveillance becomes accepted
  • •Mission creep: Border tech expands to general population

---

Legal Framework and Challenges {#legal}

International Law Obligations

States using AI at borders must comply with:

1. Refugee Convention (1951)
  • •Non-refoulement principle
  • •Right to seek asylum
  • •Non-discrimination
2. International Covenant on Civil and Political Rights
  • •Right to life (Article 6)
  • •Freedom from torture (Article 7)
  • •Right to liberty (Article 9)
  • •Right to privacy (Article 17)
3. European Convention on Human Rights
  • •Prohibition of collective expulsion (Protocol 4)
  • •Right to effective remedy (Article 13)
  • •Non-discrimination (Article 14)

Legal Challenges Underway

| Case | Forum | Issue | |------|-------|-------| | Frontex accountability | EU Courts | Pushback involvement | | Eurodac data retention | CJEU | Data protection | | Automated decision-making | National courts | Due process | | Biometric collection | ECtHR | Privacy rights | | AI asylum screening | Administrative tribunals | Fairness |

Key Legal Arguments

Against AI Border Systems:

  • •Lack of individual assessment: Violates asylum procedures
  • •Discriminatory outcomes: Violates non-discrimination principles
  • •No effective remedy: Violates right to appeal
  • •Data protection violations: Unlawful processing
  • •Disproportionate surveillance: Violates privacy rights

---

Case Studies: AI Border Failures {#case-studies}

Case 1: iBorderCtrl Lie Detector (EU)

What happened: The EU funded development of an AI "lie detector" for border screening, claiming to detect deception through facial analysis.

Problems identified:
  • •No scientific basis for technology
  • •Discriminatory outcomes by ethnicity
  • •False positive rates above 50%
  • •Violation of presumption of innocence
  • •Project abandoned after criticism

Outcome: Project discontinued, but similar technologies continue development

Case 2: Robo-Debt Australia (Parallel)

What happened: Australia's automated debt collection system wrongly accused thousands, including migrants, of welfare fraud.

Relevance to borders:
  • •Shows dangers of automated decision-making
  • •Demonstrates harm to vulnerable populations
  • •Illustrates lack of accountability
  • •Class action resulted in $1.2 billion settlement

Lesson: Automated systems require robust safeguards and accountability

Case 3: UK Visa Streaming Algorithm

What happened: UK Home Office used algorithm to "stream" visa applications, with certain nationalities automatically flagged for refusal.

Problems identified:
  • •Direct discrimination by nationality
  • •Lack of transparency
  • •No individual assessment
  • •Racist outcomes documented

Outcome: System found unlawful, discontinued after legal challenge

Case 4: US HART Biometric Failures

What happened: DHS biometric system misidentified individuals, leading to wrongful detentions.

Problems identified:
  • •Higher error rates for darker skin tones
  • •False matches with wanted persons
  • •Detention of US citizens
  • •Children misidentified as adults

Ongoing: System continues despite documented problems

---

Community Resistance and Advocacy {#resistance}

Organizations Working for Change

| Organization | Focus | Website | |--------------|-------|---------| | EDRi | European Digital Rights | edri.org | | Access Now | Global digital rights | accessnow.org | | Statewatch | EU surveillance monitoring | statewatch.org | | No Tech for Tyrants | Tech accountability | notechfortyrants.org | | Mijente | Latinx organizing + tech | mijente.net | | Collective Aid | Direct support + advocacy | collectiveaidngo.org | | Protect Not Surveil | EU AI Act campaign | protectnotsurveil.eu | | Abolish Frontex | Border agency accountability | abolishfrontex.org | | Algorithm Watch | AI accountability | algorithmwatch.org |

Successful Advocacy Campaigns

1. Amazon Rekognition Moratorium
  • •Grassroots pressure led Amazon to pause police sales
  • •Coalition of civil rights and tech groups
  • •Model for corporate accountability
2. San Francisco Facial Recognition Ban
  • •First US city to ban government facial recognition
  • •Template for other jurisdictions
  • •Shows local action can lead
3. iBorderCtrl Exposure
  • •Investigative journalism revealed program
  • •Academic research challenged claims
  • •Public pressure contributed to discontinuation

Advocacy Strategies

  • •Legal challenges: Court cases against discriminatory systems
  • •Policy advocacy: Pushing for AI Act amendments and enforcement
  • •Documentation: Recording enforcement abuses and failures
  • •Corporate campaigns: Pressuring tech companies
  • •Coalition building: Connecting migration and tech justice movements
  • •Public education: Raising awareness of surveillance harms

---

Digital Security for Migrants {#digital-security}

Protecting Yourself

Before Travel:
  • •Minimize data on devices
  • •Use encrypted messaging (Signal)
  • •Know your rights regarding device searches
  • •Consider "travel" phone with minimal data
  • •Backup and then remove sensitive contacts
At Borders:
  • •You may be required to unlock devices (varies by country)
  • •Know the rules for your destination
  • •Deletion apps may not fully erase data
  • •Cloud accounts can be accessed separately
General Digital Security:
  • •Use strong, unique passwords
  • •Enable two-factor authentication
  • •Use VPN when possible
  • •Be cautious of public WiFi
  • •Understand location tracking

Know Your Rights

At EU Borders:
  • •Right to apply for asylum
  • •Right to interpreter
  • •Right to contact UNHCR
  • •Right to legal assistance
  • •Children have special protections
At US Borders:
  • •Right to remain silent (limited at border)
  • •Right to attorney (limited at border)
  • •Citizens cannot be denied entry
  • •Device search rules vary by status

Organizations Providing Digital Security Support

  • •Access Now Digital Security Helpline
  • •EFF Surveillance Self-Defense
  • •Frontline Defenders
  • •Security Without Borders

---

What You Can Do {#action}

If You're Affected by Border AI

  • •Document everything: Screenshots, dates, names, reference numbers
  • •Request your data: Submit access requests under GDPR/privacy laws
  • •Contact legal aid: Organizations listed below provide support
  • •Report violations: To oversight bodies and human rights organizations
  • •Connect with advocates: Join campaigns for systemic change

Supporting Migration Justice

Individual Actions:
  • •Educate yourself on border technologies
  • •Support organizations listed in this guide
  • •Contact elected representatives
  • •Participate in campaigns
  • •Share accurate information
Community Actions:
  • •Organize know-your-rights workshops
  • •Support mutual aid networks
  • •Create sanctuary policies locally
  • •Document enforcement in your area
  • •Build coalitions across movements
Advocacy Priorities:
  • •Demand algorithmic transparency
  • •Push for meaningful AI oversight
  • •Protect data regardless of immigration status
  • •Require human review of AI decisions
  • •Fund independent research on border tech

---

15 Frequently Asked Questions {#faq}

1. Can AI deny my asylum claim?

AI systems are increasingly used in asylum processing, but international law requires individual assessment of each claim. However, AI may flag applications, assess credibility, or prioritize cases. Always request human review of decisions.

2. Is facial recognition used at all EU borders?

Yes, facial recognition is increasingly common at EU airports and some land borders. The Automated Border Control (ABC) gates use facial recognition, and systems are expanding.

3. What is ETIAS and how does it use AI?

ETIAS is the new EU travel authorization system launching in 2025. It uses AI to screen visa-exempt travelers against security databases and assess migration/security risks before travel.

4. Can border agents search my phone?

Rules vary by country. At US borders, agents can search devices, though recent court cases have limited this. In the EU, rules vary by member state. Know the rules before you travel.

5. What data does Frontex collect about me?

Frontex collects and processes personal data including biometrics, travel history, and information from various EU databases. You have rights under GDPR to access this data.

6. Does the EU AI Act protect migrants?

Partially. The AI Act classifies border AI as "high-risk" requiring some safeguards, but includes significant exemptions for border control and doesn't ban discriminatory systems.

7. How accurate is border facial recognition?

Accuracy varies, but studies show higher error rates for darker skin tones, women, and children. False matches have led to wrongful detentions.

8. Can I challenge an AI-based immigration decision?

Yes, but it's difficult. You have the right to human review and to challenge decisions, but getting information about how AI influenced the decision is often impossible.

9. What is a pushback and how does AI enable it?

A pushback is the illegal expulsion of people without processing asylum claims. AI enables rapid identification and processing that facilitates pushbacks before claims can be heard.

10. Are drones used to track migrants?

Yes, extensively. Both EU (via Frontex) and US (via CBP) use drones for border surveillance. Some are autonomous with AI-powered detection.

11. How can I protect my digital privacy when migrating?

Use encrypted messaging, minimize data on devices, understand your rights regarding device searches, and contact digital security organizations for guidance.

12. What organizations help challenge border AI?

Key organizations include EDRi, Access Now, Statewatch, and legal organizations in each country. See the Resources section for full list.

13. Is emotion recognition used at borders?

Yes, systems claiming to detect "deception" or "stress" through emotion recognition are being tested and deployed, despite lack of scientific basis.

14. How do I file a complaint about border AI treatment?

Contact national data protection authorities, ombudsmen, or human rights organizations. Document everything and seek legal assistance.

15. Will border AI become more invasive?

Current trends suggest yes. Budgets are increasing, technology is advancing, and regulation remains weak. Advocacy for stronger protections is crucial.

---

Resources and Organizations {#resources}

Legal Help

| Organization | Region | Website | |--------------|--------|---------| | ECRE | Europe | ecre.org | | UNHCR | Global | unhcr.org | | AIRE Centre | Europe | airecentre.org | | ACLU | US | aclu.org | | Human Rights First | US | humanrightsfirst.org | | Refugee Council | UK | refugeecouncil.org.uk |

Digital Rights Organizations

  • •EDRi: European Digital Rights - edri.org
  • •Access Now: Global digital rights - accessnow.org
  • •EFF: Electronic Frontier Foundation - eff.org
  • •Privacy International: privacyinternational.org
  • •Algorithm Watch: algorithmwatch.org

Research and Monitoring

  • •Statewatch: EU security monitoring - statewatch.org
  • •Border Violence Monitoring Network: borderviolence.eu
  • •The Migration Observatory: migrationobservatory.ox.ac.uk
  • •Mixed Migration Centre: mixedmigration.org

Advocacy Campaigns

  • •Protect Not Surveil: protectnotsurveil.eu
  • •Abolish Frontex: abolishfrontex.org
  • •No Tech for ICE: notechforice.com
  • •#NoTechForApartheid: notechforapartheid.com

Official Sources

  • •EU AI Act: ai-act-law.eu
  • •Frontex: frontex.europa.eu
  • •eu-LISA: eulisa.europa.eu
  • •UNHCR AI and Protection: unhcr.org
  • •Council of Europe: coe.int

Key Reports

  • •"AI and Migration" - Council of Europe (2024)
  • •"Inside Europe's AI Border Fortress" - IJ4EU
  • •"Protect Not Surveil" - EDRi Coalition
  • •"The Expanding Digital Border" - Privacy International
  • •"Frontex Files" - Various investigative outlets

---

Conclusion

Technology should serve human dignity, not undermine it. The fight for migration justice is fundamentally a fight for human rights in the digital age.

Key Takeaways:

  • •AI at borders is expanding rapidly with inadequate oversight
  • •Current regulations fail to protect people on the move
  • •Technology enables human rights violations at scale
  • •Community resistance and advocacy are making a difference
  • •Individual and collective action can push for change

What's Needed:

  • •Meaningful transparency in border AI systems
  • •Prohibition of discriminatory technologies
  • •Effective accountability mechanisms
  • •Real remedies for affected individuals
  • •Democratic oversight of border agencies
  • •Recognition that migration is a human right

The movement for migration justice needs technologists, lawyers, advocates, researchers, and everyday people working together to ensure that borders serve human needs, not the other way around.

---

This guide is for informational purposes and does not constitute legal advice. For specific situations, contact qualified legal assistance in your jurisdiction.

Last updated: December 2025

Tags

AI Ethics
Migration
Border Surveillance
Frontex
EU AI Act
Human Rights
Refugees
Digital Rights
Biometrics
Facial Recognition

Prepare Your Claim Privately

Organize facts, evidence, and route context before choosing your next step.