Faces in the Crowd: What Is Facial Recognition Technology — And Should It Be Banned? | Ivy Spires Knowledge Packet Issue PM-01

⚡ Ivy Spires Knowledge Packet · Issue PM-01 · June 2025

This fortnight's briefing covers one of the defining civil liberties debates of the 21st century: whether governments should be allowed to scan every face in a public crowd using artificial intelligence.

The Ivy Spires Knowledge Packet for Issue PM-01 — Faces in the Crowd — explores facial recognition technology (FRT) through the lenses of geography, history, civics, international law, and global perspectives.

Content is calibrated for both Middle School (Grades 6–8) and High School (Grades 9–12) students.

What Is Facial Recognition Technology?

Facial recognition technology (FRT) is an AI system that maps the unique geometry of a human face — the distance between eyes, the curve of the jawline, the shape of the nose — and matches it against a database of known faces. Within seconds, it can identify a person from hundreds of metres away, in a crowd, without their knowledge or consent.

Governments say it catches criminals and finds missing children. Civil rights groups say it turns every citizen into a suspect. Scientists say it makes far more mistakes on Black and Asian women than on white men — by a factor of up to 100. The question of whether to ban, regulate, or expand this technology is one of the most consequential debates of our generation.

The Numbers At A Glance

• 189 algorithms tested by US NIST in their landmark 2019 accuracy study

• 100× higher error rate for Black women vs white men in many FRT systems (NIST FRVT 2019)

• 8+ wrongful arrests in the US directly linked to FRT misidentification — all of Black men

• $14.5 billion — projected global FRT market value by 2030 (Grand View Research)

• 600 million+ cameras in China, many with live FRT integrated into a Social Credit System

• 2024: EU AI Act becomes law — the world's first comprehensive AI regulation, banning most live public FRT

  
  

How Did We Get Here? A Timeline (1960–2024)

• 1960s — FRT invented in a US defence research lab, funded by the CIA. Technology kept secret for decades.

• 2001 — First public deployment at Super Bowl XXXV in Tampa, Florida. 100,000 fans scanned. Zero confirmed matches. Called a 'mass surveillance experiment' by the ACLU.

• 2011 — Facebook's DeepFace achieves 97.35% accuracy, introducing FRT to a billion people via social media.

• 2018 — Clearview AI scrapes 3 billion+ photos from social media without consent, sells database to 600+ police agencies.

• 2019 — NIST publishes bombshell accuracy study. San Francisco becomes the first city in the world to ban government FRT.

• 2020 — Robert Williams wrongfully arrested in Detroit after FRT match error. He spends 30+ hours in detention. IBM, Amazon, Microsoft pause FRT sales to police.

• 2024 — EU AI Act bans most live public FRT by law enforcement. The US still has no federal law. China has 600 million cameras.

What Each Country Is Doing: A Global Comparison

Different countries have made radically different choices about FRT. Understanding these choices reveals fundamental differences in how nations balance security and liberty:

• 🇨🇳 China — Widespread use. 600M+ cameras integrated with the Social Credit System. Used in Xinjiang to monitor Uyghur Muslims at city scale. The state's interest in security is treated as automatically outweighing individual privacy.

• 🇪🇺 European Union — Mostly banned. EU AI Act (2024) classifies real-time public FRT by law enforcement as prohibited AI. Gold standard for democratic AI governance.

• 🇺🇸 United States — Fragmented. No federal law. Some cities ban it; federal agencies use it freely. Clearview AI contracts with 600+ police departments.

• 🇬🇧 United Kingdom — Active and contested. Metropolitan Police runs live FRT operations. Civil liberties groups challenge each deployment. No specific national law permits or prohibits it.

• 🇮🇳 India — Expanding. AFRS deployed across 16+ states. Used at Delhi airport and Kumbh Mela. No national law specifically governing public FRT.

• 🇷🇺 Russia — State security tool. Used to identify and arrest anti-war protesters in 2022 within hours of their participation in demonstrations.

Civics Connection: What Rights Are At Stake?

• Right to Privacy — UDHR Article 12 and ICCPR Article 17 protect individuals from arbitrary interference with their privacy. The UN Human Rights Committee argues FRT in public spaces without safeguards violates Article 17.

• The Chilling Effect — When people know FRT cameras are scanning them at protests or political demonstrations, many simply leave — without being arrested. Surveillance silences free expression without requiring a single prosecution.

• The Illinois BIPA (2008) — One of the world's strongest biometric privacy laws: requires written consent before FRT data collection, with $1,000–$5,000 penalties per violation. Facebook paid $650M, Google paid $100M in BIPA settlements.

• UN Call for Moratorium (2021) — The UN High Commissioner for Human Rights called for a global moratorium on public FRT, stating it poses 'unacceptably high risks to a range of human rights.'

Two Real-World Case Studies

Case Study 1: Robert Williams, Detroit USA (2020)

Robert Williams, a 42-year-old Black man, was arrested in front of his wife and daughters and held for 30+ hours in a Detroit police station. The only evidence: a facial recognition match that was wrong. He was accused of a shoplifting crime he had nothing to do with. After the error was discovered, charges were dropped — but only after Williams spent over a day in detention. He said: 'I just showed them my face. I said, I hope you don't think all Black men look alike.' He is one of at least 8 individuals — all Black men — wrongfully arrested in the United States due to FRT misidentification.

Case Study 2: London Metropolitan Police (2020–Present)

The Metropolitan Police has been deploying live FRT vans at public locations across London since 2020, scanning the faces of everyone who walks past against a watchlist of wanted persons. By 2024, these operations led to 450+ arrests. However, civil liberties group Big Brother Watch documented misidentification cases including a 14-year-old Black schoolboy stopped and questioned without reasonable suspicion. Every innocent person who walks past a van is scanned. No UK law specifically authorises or prohibits the practice — a governance gap that courts and Parliament have yet to resolve.

Key Terms Every Student Should Know

• Facial Recognition Technology (FRT) — An AI system that identifies individuals by mapping unique facial geometry and comparing it to a database.

• Biometric Data — Personal information physically unique to you (face, fingerprint, iris, DNA) that cannot be changed if stolen.

• Algorithmic Bias — Systematic errors in AI that produce unfair outcomes for certain demographic groups, typically caused by biased training data.

• Chilling Effect — The suppression of legal behaviour (protests, assembly) caused by the fear of being watched, without any direct prohibition.

• Moratorium — A temporary pause on an activity while rules are developed. Different from a permanent ban.

• The Brussels Effect — The mechanism by which EU regulatory standards become de facto global standards because multinationals cannot maintain separate technical systems for different markets.

Discussion Questions for Students and Families

• If a machine made a mistake and a human officer confirmed it, who is responsible for Robert Williams' wrongful arrest?

• China says FRT makes cities safer. Germany says it reminds people of the Stasi and makes them less free. Can both be true simultaneously?

• What is the difference between a security guard watching a crowd and a computer scanning and storing every face? Does the difference matter?

• The UN has called for a moratorium but cannot force any country to comply. If international organisations cannot enforce decisions, do they still matter?

About the Ivy Spires Knowledge Packet

The Ivy Spires Knowledge Packet is a premium fortnightly knowledge briefing for middle and high school students aged 12–17.

Each issue delivers exhaustive, curriculum-connected analysis on one global topic — spanning geography, history, international relations, civics, ethics, and MUN preparation — in two age-calibrated editions: Middle School (Grades 6–8) and High School (Grades 9–12).

This free blog post is the digest version of Issue PM-01.

The complete paid packet includes:

• a Middle School Edition with Fast Facts grid, illustrated historical timeline, four-country perspective analysis, civics connection, two full case studies, discussion questions, a letter-writing activity, and a six-term glossary; and

• a High School Edition with analytical overview using advanced terminology (algorithmic bias, hydrological coercion, the Brussels Effect), the NIST data deep-dive, a six-country comparison table with legal frameworks, complete international law analysis (ICCPR Art.17 / EU AI Act / GDPR / Illinois BIPA), MUN preparation with HRC bloc positions and operative clause templates, four university-application-depth critical analysis questions, and a model UN resolution drafting activity.

Subscribe for Rs. 999 per fortnight at ivyspires.com/subscribe.