JUTZI Philipp

Mass surveillance: No – automatic number-plate recognition (ANPR)

Mass surveillance: No

Automatic number-plate recognition – scanning every plate without cause – has no place in a free state under the rule of law.

On 26 June 2026 the Legal Affairs Committee of the Council of States (RK-S) voted yes to a national legal basis for automatic number-plate recognition (7:3 with 1 abstention). The trigger was a Federal Supreme Court ruling of October 2024 that struck down the corresponding provisions of the Lucerne police act.

The core finding of my analysis of around 60 states: mass surveillance correlates with authoritarian versus libertarian – not with left versus right. Switzerland today stands at score 4 (right-libertarian quadrant), held back by the Federal Supreme Court. The current bill decides whether we move towards hit-only (DE/NO/CA, score 2–3) or towards mass storage (NL/UK/FR, score 6). Hit-only or not at all.

→ Full dossier as PDF (7 pages, PhoneRead)

OL-DE

1 · The facts in Switzerland

What it is about. Automatic number-plate recognition (ANPR) uses cameras to capture the plates of passing vehicles and matches them automatically against police databases. The Lucerne model additionally provided for capturing the occupants and for algorithmic analysis including facial recognition. At federal level the FOCBS has operated an ANPR system with around 400 cameras since 2005, based on customs law.

The trigger. In October 2024 the Federal Supreme Court struck down the corresponding provisions of the new Lucerne police act: the cantons lack the legislative competence for surveillance measures for the purpose of criminal prosecution; moreover the purpose of use and data handling were insufficiently regulated and the interference with the fundamental rights of uninvolved persons too far-reaching.

The positions. Supporters (including the SVP) demand a national legal basis so that the police can use the tool with legal certainty – citing the mobility of offenders and the prevention of serious threats. Opponents (left-wing parties) see the suspicionless, blanket capture as a disproportionate interference with the fundamental rights of uninvolved citizens.

The names. Barbara Steinemann (SVP/ZH, National Councillor) submitted the parliamentary initiative at federal level; the Lucerne SVP section launched the cantonal attempt; the Lucerne government supported the motion.

State of proceedings. The National Council committee approved the initiative. On 26 June 2026 the RK-S also agreed (7:3 with 1 abstention). The National Council committee can now draft a bill, probably coordinated with the announced partial revision of the Federal Constitution and the amendment of the BPI. In the end the people and the cantons are likely to decide.

Transparency gap. How the individual RK-S members voted is not public – in Switzerland committee votes are not published by name. Binding decisions are taken in plenary (by name) and, for a constitutional amendment, at the ballot box. The gap concerns the committee stage.

OL-DE

2 · Scoring logic (severity score)

Combined value 1–10 = invasiveness + weakness of safeguards.

Main factor data retention: hit-only (non-hits deleted immediately) → low. Mass storage (all plates for weeks/years) → high. This axis separates a 3 from a 6.

Aggravating: no legal basis/oversight; scope creep (facial recognition, occupants, social credit); use against minorities/dissidents; centralisation. Mitigating: purpose limitation; independent data-protection supervision; judicial authorisation.

Ceiling: functioning democracy with courts → max. ~6. Authoritarian state without brakes → default 9–10.

OL-DE

3 · Political matrix

Axes. X = economy: −10 (left) … +10 (right). Y = authority: −10 (libertarian) … +10 (authoritarian).

Quadrants. Q1 = right-authoritarian (X+/Y+) · Q2 = left-authoritarian (X−/Y+) · Q3 = left-libertarian (X−/Y−) · Q4 = right-libertarian (X+/Y−).

Format per country: [quadrant | X/Y | word label]. Verified June 2026.

OL-DE

4 · Ranking (worst first)

Score 10 – Total state surveillance

China .......... 10 | Q2 | X−6 Y+10 | left-authoritarian (CCP, Xi)
North Korea .... 10 | Q2 | X−7 Y+10 | left-totalitarian (Kim)
Turkmenistan ... 10 | Q1 | X+2 Y+10 | personalist-authoritarian
Eritrea ........ 10 | Q2 | X−4 Y+10 | one-party authoritarian

Score 9 – Authoritarian mass surveillance

Russia ......... 9 | Q1 | X+5 Y+9  | right-nationalist authoritarian
Iran ........... 9 | Q1 | X+3 Y+10 | religious-authoritarian
UAE/Dubai ...... 9 | Q1 | X+7 Y+9  | monarchical-authoritarian
Saudi Arabia ... 9 | Q1 | X+6 Y+10 | religious-monarch. authoritarian
Belarus ........ 9 | Q2 | X−3 Y+9  | left-authoritarian
Egypt .......... 9 | Q1 | X+4 Y+9  | military-authoritarian
Azerbaijan ..... 9 | Q1 | X+4 Y+9  | dynastic-authoritarian
Uzbekistan ..... 9 | Q1 | X+2 Y+9  | presidential-authoritarian

Score 8 – Broad use, hardly any rules

India .......... 8 | Q1 | X+4 Y+6  | right-nationalist (BJP)
Singapore ...... 8 | Q1 | X+6 Y+7  | right-technocratic (PAP)
Vietnam ........ 8 | Q2 | X−5 Y+8  | left-authoritarian (CPV)
Myanmar ........ 8 | Q1 | X+3 Y+10 | military junta
Tuerkiye ........ 8 | Q1 | X+4 Y+7  | right-Islamic-conservative
Pakistan ....... 8 | Q1 | X+3 Y+7  | military-shaped hybrid
Kazakhstan ..... 8 | Q1 | X+3 Y+8  | presidential-authoritarian

Score 7 – Broad use / weak limits

Thailand ....... 7 | Q1 | X+3 Y+6  | royalist-conservative
Indonesia ...... 7 | Q1 | X+3 Y+5  | right-nationalist
Malaysia ....... 7 | Q1 | X+1 Y+4  | centrist coalition
Cambodia ....... 7 | Q1 | X+2 Y+9  | dynastic-authoritarian
Laos ........... 7 | Q2 | X−5 Y+8  | left-authoritarian
Mexico ......... 7 | Q3 | X−4 Y+3  | left-populist (Morena)
Brazil ......... 7 | Q3 | X−3 Y+2  | centre-left (Lula)
South Africa ... 7 | Q3 | X−3 Y+3  | centre-left (ANC)
USA (overall) .. 7 | Q1 | X+5 Y+5  | federally inconsistent

Score 6 – Democracies with mass storage

United Kingdom   6 | Q3 | X−2 Y+2  | centre-left (Labour)
Netherlands .... 6 | Q1 | X+3 Y+3  | centre-right
France ......... 6 | Q4 | X+2 Y+2  | liberal centre
Denmark ........ 6 | Q3 | X−2 Y+3  | centre-left
Hungary ........ 6 | Q1 | X+4 Y+6  | right-nationalist (Fidesz)
Israel ......... 6 | Q1 | X+4 Y+6  | right-religious

Score 4–5 – EU framework / moderate

Belgium ........ 5 | Q1 | X+2 Y+2  | centre-right
Poland ......... 5 | Q4 | X+2 Y+2  | liberal centre (Tusk)
Italy .......... 5 | Q1 | X+4 Y+4  | right-conservative (Meloni)
Spain .......... 5 | Q3 | X−3 Y+1  | centre-left (PSOE)
Australia ...... 5 | Q3 | X−2 Y+2  | centre-left
South Korea .... 5 | Q1 | X+2 Y+4  | presidential democracy
Austria ........ 4 | Q1 | X+2 Y+3  | centre-right
Japan .......... 4 | Q1 | X+3 Y+3  | right-conservative (LDP)
Czechia ........ 4 | Q1 | X+2 Y+3  | centre-right
Ireland ........ 4 | Q4 | X+1 Y+1  | liberal centre
Portugal ....... 4 | Q1 | X+2 Y+2  | centre-right
Switzerland(FOCBS) 4 | Q4 | X+3 Y+1  | liberal-federal [FSC 2024 = brake]

Score 1–3 – Hit-only / strong limits

Estonia ........ 3 | Q4 | X+3 Y+0  | liberal free-market
Finland ........ 3 | Q3 | X−2 Y−1  | strong data-protection tradition
New Zealand .... 3 | Q3 | X−1 Y−1  | centre-left, community-based
Norway ......... 3 | Q3 | X−3 Y−1  | immediate deletion
Germany ........ 3 | Q1 | X+2 Y+2  | NO storage of non-hits
Canada ......... 2 | Q3 | X−2 Y−1  | strictest hit-only
OL-DE

5 · Correlation analysis

What does NOT correlate. The economic left-right axis (X): no correlation. China (state-socialist, X−6) and Saudi Arabia/UAE (capitalist-monarchist, X+6) both reach 9–10. Wealth/GDP does not correlate either – rich (UAE, Singapore) and poor (Eritrea, Myanmar) states are equally high.

What correlates (strongest predictor first). Concentration of power (one-party states, absolute monarchies, personalist presidential systems) – the fewer independent counterweights, the higher the score. Lack of independent judicial control – the dividing line between a 6 and a 9 is whether a court can strike the system down. Weak/absent data-protection regime. State ideology of "security over individual rights", whether packaged as communism, nationalism or religious rule.

Core finding. The severity score follows the Y axis (authority) almost entirely, not the X axis (economy). Every country at Y+8 or above reaches 9–10; every country below Y±0 sits at 2–3. Both left-authoritarian (China) and right-authoritarian (Saudi Arabia) states end up at the top; both centre-left (Canada) and liberal-federal (Switzerland) states at the bottom.

In one line: it is a question of authoritarian vs. libertarian – NOT of left vs. right.

Where Switzerland stands. Q4, X+3/Y+1, score 4. Near-libertarian range with a functioning brake (Federal Supreme Court 2024). The current bill decides whether Switzerland slides up the Y axis (mass storage NL/UK, score 6) or stays down (hit-only DE/NO/CA, score 2–3).

OL-DE

4b · USA: one country, the whole spectrum

"USA overall = 7" is a rough average. In reality there is no federal rule – each state (and often each individual agency) decides its own retention period. The range runs from de-facto hit-only to years of mass storage. Exactly the same hit-only-versus-mass-storage axis as between countries – only inside a single state.

De-facto hit-only (very short retention)

New Hampshire .. 3 minutes   | non-hits deleted at once → score ~2
Utah ........... 14 days     | very short → score ~2–3
Maine .......... 21 days     | police/public safety only → score ~3
Virginia ....... 21 days     | set in law since 2025

Short to medium

Georgia ........ 30 days
California ..... 60 days     | CHP; sale to non-police prohibited
Minnesota ...... 60 days     | mandatory policy + audits
Montana ........ 90 days
North Carolina . 90 days     | audit and access obligations
Tennessee ...... 90 days

Long to mass storage

Arkansas ....... 150 days
Nebraska ....... 180 days
Vermont ........ 18 months
Colorado ....... 3 years
Florida ........ 3 years     | data confidential (no public record)
New Jersey ..... 3 years     | by AG directive
Alabama ........ 5 years     | longest statutory period

The rest without a law: in more than 30 states there is no statewide limit at all – the retention period is set by the individual agency. That goes as far as "unlimited": in a nationwide survey 13% of agencies said they keep ALPR data indefinitely. Example San José (CA): only after lawsuits in 2025/26 was retention cut from 1 year to 30 days, with a ban on capture outside churches and health clinics.

Lesson for Switzerland: without a hard statutory deletion deadline in the text of the law, a system almost always drifts towards mass storage over time. The deadline belongs in the law – not in the discretion of the agency.

OL-DE

Note on data and method

Political coordinates are rounded expert assessments, not measured indices; borderline cases may vary by ±2 points. The correlation statements rest on reading the pattern in this table, not on a computed coefficient. Government labels verified June 2026.

OL-DE

TS120

Fact check5/10

The Swiss facts are sourced. 12 of 52 countries carry a quadrant that contradicts the page's own axis definition, Switzerland included. The core finding about Y+8 is contradicted by 5 countries in the page's own table.

As of 21.09.2026