Written evidence submitted by Obinna Patrick, Clever Solution AI (WRP 25)
Electronic Voting in the House of Commons: A Trust-by-Design Framework for Secure, Auditable and Accessible Parliamentary Divisions
Electronic voting should not be treated as a convenience upgrade. It should be treated as a constitutional trust system. The question is not simply whether MPs can vote faster, but whether any changed division process would preserve legitimacy, authentication, accountability, auditability, resilience, accessibility and public confidence. This submission recommends cautious, conditional support for a limited Estate-only pilot, but not full remote voting unless further evidence demonstrates that authentication, coercion resistance, availability and public confidence risks can be mitigated. The current system already contains technology: MPs tap security passes in the lobbies while tellers provide the definitive count. That hybrid position creates useful lessons, including the need for a single authoritative audit trail. A safe pilot should begin as a shadow exercise alongside the current binding procedure, be independently audited, retain a manual fallback, publish non-sensitive performance data, and include an accessibility review. The Committee’s central question should be: what minimum trust architecture must exist before any digitised division process is legitimate?
A division is not merely an administrative count. It is the formal mechanism by which the House records political decision, personal accountability and institutional authority. Any reform to divisions therefore carries more than operational significance.
The current Commons process already combines physical procedure and technology. MPs pass through the division lobbies and tap security passes to record their names; the tellers’ count remains the definitive result. Official guidance also recognises that discrepancies can sometimes occur between teller totals and pass-reader name lists. [S7] That existing hybrid reality is the proper starting point: the question is not technology versus tradition, but what form of technology preserves trust.
This submission proposes a Parliamentary Trust Architecture Framework. It is not an argument for permanent electronic voting by default. It is a decision framework for distinguishing safe, auditable and procedurally legitimate models from weak or premature models.
The Committee should treat at least five models separately: current lobby voting, Estate-only electronic voting, chamber-proximate secure voting stations, remote electronic voting, and proxy voting. They are not different flavours of the same reform; they alter different trust conditions.
Current lobby voting combines physical presence, witnessed movement through the lobbies, teller counting and electronic name recording. That model gives visible procedural assurance but can be slow and can create accessibility barriers.
Estate-only electronic voting would preserve the requirement that MPs are on the Parliamentary Estate while changing how votes are recorded and counted. Its main advantage is that Parliament can control more of the infrastructure, equipment, network environment, staffing and immediate dispute-resolution process.
Remote electronic voting changes the risk profile more fundamentally. The MP may be outside the Estate, using networks and devices outside parliamentary control. Surrey Centre for Cyber Security evidence to this inquiry identifies remote voting as more difficult to mitigate than Estate voting because more remote hardware, software and network elements must be defended. [S17]
Proxy voting is different again. It accepts absence but preserves a known human intermediary operating under approved rules. It should not be treated as equivalent to a remote electronic vote cast directly by the absent Member.
Perception is not cosmetic. In parliamentary voting, perceived legitimacy is part of legitimacy. If the public, Members or journalists cannot understand who voted, how the vote was recorded, how errors are corrected and which number is authoritative, confidence can erode even where the technical result is correct.
The current system has an official explanation for discrepancies: teller totals decide the result, while pass readers generate the published names and missing names may later be corrected. [S7] That is procedurally clear, but it can be difficult for the public to understand after a close or controversial vote.
An electronic system could improve perception if it produces one clear authoritative public record, but it could worsen perception if it becomes a “black box”. Therefore, the MP-only principle requires more than login credentials. It requires strong authentication, real-time confirmation to the Member, an audit log, challenge and correction rules, and a published explanation understandable to non-specialists.
The Committee should require any proposed model to show exactly how an MP can verify that their vote was cast as intended and recorded as counted.
The most relevant lesson is not that another legislature uses electronic voting, but which trust conditions accompany it. The House of Lords example is particularly relevant because the Lords page states that Members continue to vote online but may only do so when present on the Parliamentary Estate. [S12] That is not full remote voting; it is a controlled Estate-based model.
The US House example shows that electronic voting can be integrated into a legislature for recorded votes. Members have used voting cards and voting stations around the Chamber since 1973, and the official history records that the system reduced the time needed for recorded votes. [S13] However, institutional transfer is not automatic. The Commons has its own procedural culture, whipping practices, division lobbies and accountability norms.
Comparators should therefore be used as design evidence, not as shortcuts. The right question is: which model preserves the Commons’ procedural values while improving speed, accessibility and record accuracy?
The official Covid-19 proceedings page records that remote voting was introduced in the Commons on 12 May 2020, that the temporary arrangements expired on 20 May 2020, and that pass-readers continue to be used in divisions. [S11] The Committee should separate the brief remote-voting episode from the later and continuing use of pass-readers in physical divisions.
One lesson is that parliamentary voting technology must be evaluated after use. Surrey Centre for Cyber Security evidence states that there has been no published review of the performance and security of the Covid remote voting system. [S17] This submission cannot independently verify that no such review exists; therefore, the safe recommendation is that the Committee should ask the House Service and Parliamentary Digital Service to identify any internal reviews, incident logs, user research, security assessments and lessons-learned papers that can safely be disclosed.
A second lesson is that emergency procedures should not automatically become permanent procedures. A pandemic can justify temporary risk acceptance. Permanent reform requires a higher standard of evidence, testing and governance.
Electronic voting could affect more than the mechanics of counting. It could alter the relationship between debate and decision, the timing of divisions, the role of tellers, the visibility of Members, access to Ministers in the lobbies, the management of multiple divisions, correction procedures, publication of results, pairing, proxy voting and expectations of physical attendance.
HM Government evidence to this inquiry argues that many Members see lobby voting as an opportunity to engage senior colleagues, including Ministers, and that votes and debates should be kept together. [S16] Whether or not the Committee accepts that position, it identifies a real procedural question: should electronic voting preserve the deliberative moment around the vote, or should it separate voting from the physical Commons timetable?
Any recommendation should therefore state explicitly which surrounding practices would change and which would remain unchanged. A technical pilot without procedural boundary conditions would be a weak pilot.
The House Administration would need to support authentication, user training, equipment management, service availability, cyber assurance, accessibility testing, incident response, data governance, publication systems and post-division correction procedures.
The Parliamentary Digital Service provides digital services to Parliament and has responsibilities including cyber security, the parliamentary network and software supporting parliamentary work. [S14] That makes PDS central to any serious implementation, but this submission has not found verified evidence of the exact staff, cost or procurement requirements for a new Commons electronic voting pilot. INSufficient evidence should not become pretend precision.
The Committee should ask for a costed implementation note covering staffing, assurance, procurement, audit, support cover during sittings, accessibility support and contingency operations.
The accessibility question is substantial. The present system depends on physical movement through the lobbies within a time window. Erskine May records that if Members disabled by infirmity are present on the Estate and wish to vote, their names may be communicated by a Whip to the Clerks and tellers and included in the count. [S10] That shows the current system already recognises exceptional physical barriers.
Estate-only electronic voting may improve accessibility for some Members if it reduces queuing, congestion, mobility difficulty or the need to move through a particular lobby. Secure designated voting stations may help further if they are genuinely accessible and located sensibly.
Remote voting may offer greater participation gains for illness, pregnancy, caring responsibilities or mobility barriers, but it also raises stronger authentication, coercion, availability and presence questions. Accessibility gains must therefore be measured rather than asserted. The Committee should ask what accessibility problem each model solves, for whom, and with what residual risk.
The central security concerns are integrity of the vote, integrity of the tally, availability of the system, authentication of the Member and trustworthy dispute resolution. Confidentiality is less central than in statutory elections because MPs’ division votes are public, a point also made by Surrey Centre evidence. [S17]
Mitigations should include multi-factor authentication where appropriate, Estate-controlled voting devices or terminals, separation from the open internet where possible, real-time Member confirmation, independent security testing, monitored audit logs, clear challenge windows, redundancy, manual fallback and post-division reconciliation.
Security transparency must be balanced. A 2020 House of Commons FOI response withheld the supplier of voting terminals on national security and law-enforcement grounds and recognised risks of cyberattack and unlawful interference with voting equipment. [S15] Therefore, Parliament should publish enough to build confidence in governance and auditability without publishing operational details that increase attack surface.
The Committee should test every model against the following eight tests. The framework is designed to prevent the debate collapsing into a false binary between “modern efficiency” and “tradition”.
Test | Question for the Committee | Minimum condition |
1. Legitimacy Test | Would Members and the public still regard the division result as authoritative after a controversial vote? | Requires a clear legal/procedural basis, visible accountability and an authoritative published result. |
2. Authentication Test | Can Parliament verify that the correct MP cast the vote? | Requires strong identity assurance, controlled credentials, device/terminal controls and immediate vote confirmation. |
3. Presence Test | Does the model preserve the constitutional meaning of presence, participation and accountability? | Estate-only voting preserves presence more strongly than remote voting; proxy voting preserves accountability differently. |
4. Coercion Resistance Test | Can the system reduce pressure, proxy misuse, informal whipping abuse or inappropriate remote influence? | Requires consideration of where the Member votes, who may be near them, and how dispute or undue influence concerns are raised. |
5. Auditability Test | Can votes be checked, challenged, corrected, recorded and historically audited? | Requires logs, reconciliation, challenge windows, correction rules and publication architecture. |
6. Resilience Test | What happens if the system fails during a live division? | Requires redundancy, manual fallback, rehearsed incident procedure and criteria for invalidating or rerunning a division. |
7. Accessibility Test | Does the model improve participation without creating unequal procedural advantage? | Requires user research, accessible hardware/software, physical accessibility, training and evaluation data. |
8. Public Confidence Test | Can the system be explained simply enough that people trust the result? | Requires public-facing explanation, non-sensitive audit reporting and clear correction procedures. |
Model | Benefits | Risks | Conditions required | Evidence strength |
Current lobby voting with pass readers | High visibility; physical presence; tellers provide definitive count; pass readers support publication of names. | Slow; access barriers; possible discrepancies between teller totals and pass-reader name lists; queues and congestion. | Maintain clear public explanation of teller count versus name list; improve correction reporting; accessibility review. | High for current procedure; official sources confirm operation. [S7] [S8] [S10] |
Estate-only electronic voting | Preserves Estate presence; Parliament controls more infrastructure; faster counting; easier support and dispute handling. | May reduce lobby contact; system failure could disrupt divisions; public may distrust black-box counting. | Independent audit; Estate-controlled terminals; real-time Member verification; manual fallback; public reporting. | Medium/high: supported by Lords comparator and expert evidence, but Commons-specific pilot evidence needed. [S12] [S17] |
Secure designated voting stations near Chamber | Preserves proximity to debate; reduces congestion; accessible station design possible. | Station placement could create procedural inequity; queues may merely move; still needs strong authentication. | Accessible station map; capacity testing; teller/manual fallback; clear rules on when and where voting is valid. | Medium: plausible but Commons-specific evidence needed. |
Remote electronic voting | Maximum flexibility for illness, caring, travel and emergency situations. | Harder authentication; more attack surface; weaker environmental control; coercion and availability risk; presence concerns. | Only after strong evidence; controlled devices; MFA; challenge protocol; emergency conditions; independent security review. | Low/contested for permanent Commons use. Surrey evidence says remote risks are more significant and harder to mitigate. [S17] |
Emergency-only remote voting | Preserves continuity during extraordinary disruption; avoids normalising remote voting for ordinary business. | Emergency definitions may expand; systems unused in normal times can fail under pressure; trust may suffer if deployed without rehearsal. | Clear trigger conditions; rehearsals; sunset clauses; post-event review; default return to normal procedure. | Medium: Covid proves temporary adaptation occurred, not that permanent readiness is established. [S11] |
Proxy voting | Supports absence under approved rules; retains known human procedure; already exists in Commons practice. | Does not give the absent Member direct participation; depends on proxy accuracy and party/Member arrangements. | Clear eligibility rules; transparent records; periodic review; interaction with e-voting clarified. | High for existence; separate from electronic voting. [S11] [S16] |
Risk | Why it matters | Mitigation | Evidence status | Residual concern |
Constitutional risk | A division is a constitutional act, not merely a transaction. | Define before pilot what presence, personal accountability and authority mean under each model. | Official procedure verified; constitutional implications are reasoned inference from procedure. | Residual concern: any change may alter conventions not fully captured in technical testing. |
Cybersecurity risk | Interference, disruption or credible allegation of interference could damage confidence in parliamentary decisions. | Independent security assessment; hardening; monitored logs; red-team testing; controlled Estate infrastructure; fallback. | Supported by PDS role, FOI risk language, Surrey evidence and IET general cyber evidence. [S14] [S15] [S17] [S21] | Mitigation strong for Estate-only if properly implemented; partial/unproven for remote. |
Authentication risk | The MP-only principle fails if someone else can cast or alter a vote. | MFA or equivalent; physical pass/token controls; terminal confirmation with Member name/photo; challenge window. | Current pass-reader operation verified; future design unverified. [S9] | Strong for controlled Estate model; partial for remote. |
Coercion risk | Remote or poorly controlled voting could expose Members to inappropriate pressure or disputed influence. | Prefer Estate-only model; require confidential reporting channel; avoid unsupervised remote voting unless evidence supports safeguards. | INSUFFICIENT EVIDENCE for Commons-specific coercion incidents. Risk is reasoned from remote-voting environment. | Mitigation partial. |
Procedural legitimacy risk | A technically correct system may still be unacceptable if it breaks the link between debate, presence and decision. | Define procedural boundary conditions before technical build; preserve debate-vote relationship unless House explicitly decides otherwise. | Supported by HM Government evidence and official procedure. [S16] | Residual concern depends on Member consensus. |
Public trust risk | Public confusion can become a legitimacy problem after close or controversial votes. | Plain-English public explanation; authoritative result architecture; publish correction rules and non-sensitive audit data. | Official page confirms current discrepancy mechanism. [S7] | Mitigation partial; perception cannot be engineered by system design alone. |
Accessibility risk | A reform may fail if it improves speed but not access for Members facing mobility, disability, health, pregnancy or caring barriers. | Accessibility review; user testing; accessible station design; metrics before and after pilot. | Inquiry asks protected-characteristics question; Erskine May verifies current provision for infirm Members on Estate. [S10] [S18] | Mitigation partial until tested with affected Members. |
Operational failure risk | A live division cannot be left unresolved because a system freezes, loses connectivity or records incorrectly. | Manual fallback; parallel shadow pilot; incident runbook; rehearsals; criteria for pausing or reverting. | Current teller fallback verified; future system unverified. [S7] [S8] | Strong only if fallback remains binding. |
Data protection/privacy risk | Voting records are public, but authentication logs, device data, access records and support records may contain sensitive operational data. | Data minimisation; retention rules; access controls; DPIA; clear separation of public vote record and security logs. | INSUFFICIENT EVIDENCE on proposed data architecture. Risk is reasoned. | Mitigation partial until design exists. |
Equality and participation risk | A system could create unequal advantage if some Members can vote more easily than others or if support is uneven. | Equal access, training, support, equipment standards and published accessibility evaluation. | Question is within inquiry scope; exact impact requires Member evidence. [S18] [S19] | Mitigation partial. |
Media/political controversy risk | Allegations of “app voting”, hacking or absent MPs voting from inappropriate locations could dominate coverage. | Avoid remote model at first; use Estate-only terminology; publish safeguards; no hype. | INSUFFICIENT EVIDENCE for future media reaction; risk is foreseeable but not measurable from reviewed sources. | Mitigation partial. |
Precedent risk | A limited emergency or accessibility measure may become the basis for broader reform without equivalent scrutiny. | Sunset clauses; post-pilot evaluation; House approval before permanence; separate votes for separate models. | Reasoned inference from procedural reform risk. | Mitigation strong if governance is explicit. |
The safest pilot is not immediate full adoption. It is a staged, trust-preserving pilot that allows Parliament to learn without making the technology legally decisive before it has been proven.
Stage | Design | Reason |
Stage 0: Pre-pilot assurance | Publish the model under consideration; confirm whether it is Estate-only, chamber-proximate, remote or emergency-only; commission security, accessibility and procedural impact assessments. | Prevents vague reform by label. |
Stage 1: Shadow pilot | Run electronic recording alongside current binding division procedure. The teller result remains authoritative. Compare speed, error rate, Member experience, accessibility and reconciliation outcomes. | Allows learning without risking a binding division result. |
Stage 2: Limited Estate-only live pilot | Only if Stage 1 succeeds, use the electronic system for a defined period and limited scope, with manual fallback and immediate reversal conditions. | Tests real operation while limiting constitutional exposure. |
Stage 3: Independent evaluation | Publish a report covering non-sensitive performance data, accessibility findings, incident counts, correction requests, Member feedback and cost/resourcing implications. | Creates evidence before permanence. |
Stage 4: House decision | No permanent adoption unless the House has the evaluation, the Committee reports, and Members vote on a clearly defined model. | Protects parliamentary control. |
Minimum pilot safeguards:
1. What precise problem is electronic voting intended to solve: time, accuracy, accessibility, resilience, public transparency or all of these?
2. Which model is being considered: current lobby/pass-reader improvement, Estate-only electronic voting, chamber-proximate voting stations, remote voting, proxy voting reform or emergency-only remote voting?
3. What procedural value of lobby voting must be preserved, and what value is Parliament willing to change?
4. How will the system verify that the correct MP cast the vote?
5. How will an MP confirm that their vote was cast as intended and recorded as counted?
6. What is the authoritative result if electronic totals, published names and any manual count disagree?
7. What is the manual fallback if the system fails during a live division?
8. Who will independently audit the system before and after any pilot?
9. What information will be published after each division, and what security information will remain confidential?
10. How will accessibility gains be measured rather than merely asserted?
11. What data will be collected, who can access it, and how long will it be retained?
12. What threshold must be met before any pilot becomes permanent, and what failure triggers immediate reversion?
The Committee should recommend cautious, conditional support for an Estate-only electronic voting pilot, beginning with a non-binding shadow pilot and moving to a limited live pilot only if authentication, auditability, accessibility, resilience, transparency and public confidence safeguards are satisfied.
The Committee should not recommend general remote electronic voting on the evidence reviewed here. Remote voting may be justified under defined emergency or exceptional accessibility conditions, but it should not be treated as a routine extension of Estate voting unless Parliament has stronger evidence on authentication, availability, coercion resistance and public confidence.
Electronic voting may be worth piloting, but only as a trust-preserving reform, not as a convenience-driven shortcut.
June 2026
[S1] UK Parliament Committees, Electronic voting inquiry page. https://committees.parliament.uk/work/8885/electronic-voting/
[S2] UK Parliament Committees, New inquiry launched into electronic voting in the House of Commons. https://committees.parliament.uk/committee/126/procedure-committee/news/205131/new-inquiry-launched-into-electronic-voting-in-the-house-of-commons/
[S3] UK Parliament Committees, Procedure Committee page. https://committees.parliament.uk/committee/126/procedure-committee/
[S4] UK Parliament Committees, Electronic voting inquiry publications page. https://committees.parliament.uk/work/8885/electronic-voting/publications/
[S5] UK Parliament Committees, Written evidence list for Electronic voting inquiry. https://committees.parliament.uk/work/8885/electronic-voting/publications/written-evidence/
[S6] UK Parliament, Give evidence to a select committee. https://www.parliament.uk/get-involved/committees/give-evidence-to-a-select-committee/
[S7] UK Parliament, Divisions. https://www.parliament.uk/about/how/business/divisions/
[S8] MPs' Guide to Procedure, What happens in the Chamber when there's a vote. https://guidetoprocedure.parliament.uk/articles/B3JwLFJs/what-happens-in-the-chamber-when-theres-a-vote
[S9] MPs' Guide to Procedure, How to vote in a pass reader division. https://guidetoprocedure.parliament.uk/articles/5oP67iY3/how-to-vote-in-a-pass-reader-division
[S10] Erskine May, Procedure on a division. https://erskinemay.parliament.uk/section/4802/procedure-on-a-division
[S11] UK Parliament, COVID-19 proceedings: voting. https://www.parliament.uk/about/how/covid-19-proceedings-in-the-house-of-commons/remote-voting/
[S12] UK Parliament, Online voting in the House of Lords. https://www.parliament.uk/about/how/changes-to-lords-proceedings/online-voting-in-the-house-of-lords/
[S13] US House of Representatives: History, Art & Archives, Electronic Voting. https://history.house.gov/Exhibitions-and-Publications/Electronic-Technology/Electronic-Voting/
[S14] UK Parliament, Parliamentary Digital Service. https://www.parliament.uk/mps-lords-and-offices/offices/bicameral/parliamentary-digital-service/
[S15] UK Parliament FOI, Voting terminals (2020). https://www.parliament.uk/site-information/freedom-of-information/information-we-already-publish/house-of-commons-publication-scheme/chamber-and-committees/voting-terminals-2020/
[S16] Procedure Committee written evidence, HM Government submission (EVO 11). https://committees.parliament.uk/writtenevidence/141447/html/
[S17] Procedure Committee written evidence, Surrey Centre for Cyber Security submission (EVO 10). https://committees.parliament.uk/writtenevidence/141104/html/
[S18] Procedure Committee written evidence, John Simmons submission (EVO 14). https://committees.parliament.uk/writtenevidence/142821/html/
[S19] Procedure Committee written evidence, Callie Cummins submission (EVO 21). https://committees.parliament.uk/writtenevidence/163163/html/
[S20] Procedure Committee written evidence, Centenary Action submission (EVO 12). https://committees.parliament.uk/writtenevidence/141087/html/
[S21] IET, Internet voting in the UK. https://www.theiet.org/media/9537/internet-voting-in-the-uk.pdf