SMARTMATIC PCOS 2013A Primer on the

Automated Election System

in the Philippines

By the Automated Election System Watch (AES Watch)

(What the Filipino people should know about the Smartmatic PCOS thatwas proven to be deficient with admitted program errors remaininguncorrected. The use of automation technology without the license fromthe real software owner, Dominion voting Systems, looms in the comingMay 2013 elections with the DVS terminating its licensing agreement withSmartmatic, last May 23, 2012. An ongoing legal battle between Smartmaticand DVS filed in September 2012 in the chancery Court of Delaware, USAreveals the core issue at the heart of the PCOS problem: that the electiontechnology used in May 2010 was not fully functional, had programmingerrors, with the technology owner accusing Smartmatic of notcomplying with international standards, a validation of studies by FilipinoIT experts and social scientists as early as 2009. The ongoing battle of citizens'watchdogs and the Filipino IT community for the integrity of the vote and accountability in automated elections remains integral with the fight for the Freedom of Information --the right to know- by the public.)

Launched on Jan. 18, 2010, the Automated Election System Watch (AES Watch) is now composed of more than 40 organizations, among them, the University of the Philippines Alumni Association (UPAA), National Secretariat for Social Action-CBCP, Center for People Empowerment in Governance (CenPEG), Philippine Computer Society (PCS) Foundation, TransparentElections.org, Computer Professionals Union, Caucus of DLSU-CCS, NAMFREL, Association of Major Religious Superiors of the Philippines (AMRSP), National Union of Students of the Philippines (NUSP), Philippine Computer Society Foundation; Transparency International-Philippines, National Council of Churches in the Philippines (NCCP), Faculty of Ateneo de Manila Dept. of Information Communications System, UP Dept. of Computer Science and ITTC, Philippine Computer Emergency Response, Movement for Good Governance, Concerned Citizens Movement, Dilaab-Hearts Foundation, Solidarity Philippines, Association of Schools of Public Administration in the Philippines (ASPAP), Sisters Association in Mindanao (SAMIN), Computing Society of the Philippines (CSP), Pagbabago (Movement for Social Change), Alyansa Agrikultura, Ecumenical Bishops Forum, Biliran Kawsa, Workers' Election (WE) Watch, Health Alliance for Democracy, and others. Its President Emeritus is former Vice President Teofisto Guingona, Jr. while its first spokesperson was Alfred Pascual, now UP President.

For information, please contact: Acting Secretariat Office, 304 CSWCD Bldg., Magsaysay Avenue, UP Diliman 1101 Quezon City; TelFax +9299526; email address: aeswatch2012@gmail.com

AUTOMATED ELECTION SYSTEM WATCH (AES Watch)

A Primer on the

Automated Election System

in the PhilippinesBy the Automated Election System Watch (AES Watch)

I. The automated election system (AES)

A. What is the Automated Election System (AES) for Philippine elections?

The AES for Philippine elections is mandated by Republic Act (RA) 9369. RA 9369 amends RA 8436, entitled "An Act authorizing the Commission on Elections to use an automated election system in the May 11, 1998 national or local elections and in subsequent national and local electoral exercises, to encourage transparency, credibility, fairness, and accuracy of elections, amending Batas Pambansa Blg. 881, as amended, RA No. 7166 and other related election laws, providing funds thereof and for other purposes”; “to use an automated election system or systems in the same election in different provinces, whether paper-based or a direct recording electronic election system as it may deem appropriate and practical for the process of voting, counting of votes and canvassing/consolidation and transmittal of results of electoral exercises” a system using appropriate technology which has been demonstrated in the voting, counting, consolidating, canvassing, and transmission of election results, and other electoral process.

SEC. 1 of RA 9369: It is the State policy to ensure free, orderly, honest, peaceful, credible and informed elections, plebiscites, referenda, recall and other similar electoral exercises by improving on the election process and adopting systems, which shall involve the use of an automated election system that will ensure the secrecy and sanctity of the ballot and all election, consolidation and transmission documents in order that the process shall be transparent and credible and the results fast, accurate and reective of the genuine will of the people.

1As AES Watch has always consistently advocated, voter education

and poll watching should focus not only on the external features and operations of the AES but more critically on its internal systems. Vigilant watching should as well monitor the Comelec, the foreign technology provider, and other principal agencies and implementers of poll automation as well as the election cheats. There is no substitute for vigilance and in safeguarding our freedom.

AES Watch is adopting the 12 ALERTS released by CenPEG as an aid for voters and watchdogs on Election Day. Meanwhile, AES Watch believes that all efforts at ensuring that accountability for non-compliance should be a major continuing concern of the Filipino voters and all stakeholders to relentlessly pursue to promote the integrity of the vote and transparency of the elections.

36

B. Under the poll modernization law (RA 9369), why does the country need to automate its elections?

Automating the election, lawmakers said, will eliminate clerical, human intervention-related errors. Other reasons: The conventional manual process is too long and tedious. It takes almost two months before national positions are proclaimed; to remove conditions for dagdag-bawas or wholesale cheating in manual elections.

C. What is the automated process of voting?

AES as contracted to the vendor in 2010 and again in 2013 entails the process of using the computer to electronically count, record, store, transmit and canvass the votes that come in to the machines. In the 2010 elections, the PCOS (Precinct Count Optical Scan) in the clustered precinct of five precincts with 1000 voters at most, was used to accept, count, and store in scanned images the shaded ovals of the 30 inch ballot containing more than 100 names of the candidates. After electronically counting the shaded ovals, the number of votes added by the machines is transmitted to the CCS (Consolidated Canvassing system) server for the MBOC (Municipal Board of Canvassers) to base the canvassed votes.

D. What are the requirements of a credible, secured, and reliable automated election system (AES)?

The Automated Election System Watch (AES Watch), through a committee of IT, business, and management experts came up in 2009 with its System Trustworthiness, Accountability, and Readiness (STAR) Card to comprehensively assess and rate the implementation of AES primarily by Comelec and Smartmatic-TIM. The STAR Card listed 20 items of concerns rated as Pass=3; Qualified Pass=2.5; Warning=2; Danger=1; Fail=0.

a. System set-up (Will the AES be ready for full implementation?)

Timely delivery of machines Quality of machines Technology certification Availability of transmission facilities Deployment of machines Physical security of machines

2amendments for enhancing RA 9369, held dialogs with Comelec, CAC, Congress, DOST, TEC, and other agencies, and the convening of JCOC. In cooperation with UP, it organized the first Filipino IT for Election (FIT4E) national conference in June 2011 where key IT groups, practitioners, academics, researchers, U.S.-based IT scholars, and multi-sectoral groups resolved to make election technology inclusive by tapping the expertise of Filipino IT – a resolution that was backed, at least in principle, by Comelec commissioners.

Likewise, AES Watch and affiliate organizations, for three years, pressed for Congress through its election committees to review the May 2010 election automation and probe into the accountability of the election manager and technology providers. Similarly, several bills were proposed to ensure compliance with the election law. In April 2012 individual conveners and members of AES Watch petitioned the SC for a temporary restraining order against Comelec's option to purchase the Smartmatic machines.

IV. Preparations for the May 2013 mid-term elections

A. What can we as citizens, voters, and election watch groups do for the May 2013 automated elections and for our country?

The right to suffrage is the people's sovereign right to elect officials upon whom they confer the authority to serve the public responsibly, transparently, with accountability, and on a full-time basis. For this reason, it is imperative that the electoral process is credible, trustworthy, reliable, and accurate in serving as the instrument of the people's will.

However, given the learning lessons of the first automated elections, the lack of transparency and accountability in the system, its implementers, and oversight bodies as well as the persistence of traditional fraud which automation itself has not effectively addressed – ensuring the electoral process fair, democratic, and reliable rests on the vigilance of the people.

Automating the election is fine – but it should comply with the law and the high standards of IT and election management. While we continue to push for a reliable technology, we should be alert on the fact that modern technology's capability to eliminate fraud, whether traditional or electronic, remains to be validated.

35

Precinct-specific ballots Resource inventory at voting centers Adequate general instructions

b. Internal security (Will the AES have the necessary safeguards to prevent fraud?)

Source code & its review Verifiability of voting and results Secured transmission of results Initialization of machines Random Manual Audit of vote counts

c. Personnel training and voters' education (Will the teachers and the voters know exactly what to do on election day?)

Training of election personnel Stakeholders education & training Precinct assignment voters

d. Contingency planning

Continuity plan Electoral protest mechanism Alternative election system

E. According to the law, who are responsible for making the AES successful?

Government: Comelec Project Management Office together with system-integrator Smartmatic counterparts; Comelec Advisory Council; Technical Evaluation Committee; DOST-certified IT-capable BEIs; Board of Canvassers; Joint Congressional Oversight Committee (JCOC) on AES; Armed Forces of the Philippines-Philippine National Police; and other state agencies.

Others : International Certification body; election watchers; media such as the Kapisanan ng mga Brodkaster sa Pilipinas (KBP); Partish Pastoral Council for Responsible Reporting (PPCRV), and the National Movement for Free Elections (NAMFREL); Political Parties and Candidates.

3announced recently there will be no source code review – the second time when this vital legal requirement will not be complied with.

Pressed for transparency, Comelec under its new head held dialogs with election stakeholders led by AES Watch. Contrary to claims that it welcomed strong policy recommendations to correct the Smartmatic-TIM AES program bugs and other vulnerabilities in the final analysis Comelec refused to adopt the legally-mandated minimum system requirements such as source code review and digital signature and still exercised its “option to purchase” the PCOS machines for use in 2013. Under its new membership, the Comelec Advisory Council (CAC) welcomed demands for reforming the automation system and recommended the non-purchase of the PCOS machines. Its recommendations were, unfortunately, thumbed down by Comelec.

The Supreme Court (SC) in a Sept. 21, 2010 landmark decision directed the Comelec to release the source code for review by CenPEG and other election stakeholders. The review of the AES 2010 source code was never held due to restrictions imposed by the Comelec that made an independent, scientific, and rigorous review by IT experts impossible.

As expected, the technology provider insisted that its system was sound and boasted to the whole world about its “election success” in the Philippines. Facing persistent demands from citizens' election watch groups, Congress' election-related committees held hearings on electoral reform. As a whole however the lower House endorsed the use of the Smartmatic-TIM system for the 2013 elections. Only dogged follow-ups by AES Watch made the JCOC to convene finally in November 2012. But the assessment of the 2010 poll automation was never clearly part of its agenda.

B. How were these major findings and conclusions addressed by other election stakeholders?

Under its old leadership, the accredited citizens' watchdog, PPCRV, as expected stood for the re-use of the Smartmatic-TIM system in the 2013 elections and never indicated receptiveness to proposals by various citizens groups' including its own ground-level volunteers in the provinces to either reconsider the 2010 PCOS machines or make the technology provider accountable for the errors committed.

On the other hand, since Day 1 after the 2010 elections, AES Watch and its affiliate groups along with other watchdogs lobbied for several

34

F. What is the role of the other sectors in the conduct of AES?

Other sectors/stakeholders mandated by the Constitution and other laws: the civil society community, NGOs, citizens' watchdogs, media, institutions, and the general public.

Their role: 1) right to participate in the electoral exercise as well as in policy- and decision-making; 2) they represent the sovereign will of the people and government emanates from them; 3) RA 9369 gives CSO and other “interested parties” the right to review the election source code, monitor the conduct of elections including the RMA, conduct dialogs with Comelec and related agencies, participate in Congress hearings including proposing amendments to existing laws, and file election protests; 4) to protect the right to suffrage, right to public information and transparency, non-interference by foreign entities, as well as to fair, honest, and credible elections they have the right to hold all government agencies and officials accountable for misdeeds with the actions to include going to court and calling for impeachment.

G. When was the AES first implemented?

The first AES was implemented in ARMM elections in 1996 and then in August 1998. There was an attempt to fully automate the national and local elections in 2004 but the Supreme Court stopped the implementation. The 2007 midterm elections was not automated due to lack of time. The May 10, 2010 presidential elections was the first to be automated.

H. How was the automated system conducted in 2010?

What were the activities – as well as critical problems and issues - on election day (May 10, 2010) and after? Were these problems and issues addressed by Comelec and other concerned agencies?

4irregularities were vivid in Lanao del Sur. (Testing Democracy: The 2010 Automated Elections in Lanao del Sur, Dec. 2010)

People's International Observers Mission (PIOM): “In focusing on the machines, the Comelec lost the people. The elections were not peaceful or violence-free. Neither were the elections fair nor honest.”

Global Filipino Nation (GFN, election observers' team): “…Events put to question the authenticity, integrity, confidentiality, veracity and accuracy of the vote counts in the ERs. The dark cloud rose from disabling critical, legally specified security features, particularly relating to the digital signatures. Thus, no one (both perceived winners and losers) can be sure whether the vote results are true and correct, and reect the real will of the Filipino people. Accordingly, the Election Observers Team of GFN challenges the legitimacy of the election results.”

III. Accountability

A. Are the critical issues and main problems being addressed by Comelec and other responsible agencies?

The critical issues are not being addressed properly or are largely

ignored. Just like in 2010, CF cards will be used again for 2013. The controversial machine-generated “digital signature” will be used again.

The July 24-25, 2012 “mock elections” held by Smartmatic at the lower House revealed a 97% accuracy rating which is lower than the required 99.995% rating. The low accuracy translates to 600 errors out of 20,000 ballot marks (the law mandates only 1 out of 20,000). The deficiencies and inaccuracies of the Smartmatic PCOS system were exposed again – but were belittled by Comelec - during the Feb. 2, 2013 “mock elections” which resulted in machine breakdowns, paper jams, transmission delays, and counting discrepancies.

The system that will automate the May 2013 mid-term elections has no certification by the international SysTest Labs, Inc. in the absence of approval by the technology owner, Dominion. Comelec said they will use the voting system designed for the aborted 2011 ARMM election which is regional unlike the 2013 mid-term elections with thousands of elective positions at stake. Comelec Chairman Sixto Brillantes also

33

Ph

ase

Issu

eC

om

elec

Act

ivit

y /

S

ub

-act

ivit

y

Ap

pra

isal

of

Co

mel

ec A

ctio

n

Ele

ctio

n D

ayE

lect

ion

mat

eria

ls

inv

ento

ry;

Bo

oti

ng

of

PC

OS

&

sig

nin

g i

n;

Zer

o r

esu

lts

pri

nti

ng

;D

ecla

rati

on

th

at p

oll

is

op

en f

or

vo

tin

g;

Tec

hn

ical

pro

ble

ms

of

bo

oti

ng

an

d s

ign

ing

in

by

BE

I in

sev

eral

cl

ust

ered

pre

cin

cts;

Do

cum

ente

d s

can

ned

v

ote

s al

read

y

dis

pla

yed

on

th

e P

CO

S b

efo

re s

tart

of

vo

tin

g a

s in

Bil

iran

.

2010

: T

he

syst

em w

as a

“r

eso

un

din

g s

ucc

ess!

”

2011

: Th

e sy

stem

was

a

“qu

alifi

ed s

ucc

ess!

”

No

tec

hn

ical

rep

ort

w

as d

on

e w

ith

ac

com

pan

yin

g

do

cum

ents

rel

ease

d t

o

the

pu

bli

c.

Do

cum

ents

ex

pla

inin

g

the

nat

ion

wid

e p

rob

lem

s sh

ou

ld b

e d

iscl

ose

d t

o t

he

pu

bli

c.

Co

mel

ec a

nd

S

mar

tma

tic

con

ten

d

that

th

e p

rob

lem

s w

ere

exp

ecte

d o

f a

new

sy

stem

sin

ce “

no

te

chn

olo

gy

is

per

fect

an

yw

ay.”

Mas

siv

e d

isen

fran

chis

emen

t d

ue

to 1

-bla

nk

bal

lot

Vo

tin

g p

erio

dT

he

old

5 p

reci

nct

s-p

er c

lust

ered

pre

cin

ct-

con

fig

ura

tio

n h

as n

o

Th

e k

ilo

met

ric

qu

eues

o

n e

lect

ion

day

wer

e th

e b

est

arg

um

ent

5…digital signatures in the PCOS machines, contrary to the claims of Smartmatic”; “the PCOS machine contains a console port”; (the Team upon testing) “was able to connect an ordinary laptop computer to the console port of a PCOS machine”. (The Joint Forensic Team, led by Atty. Al Vitangcol III, examined 60 PCOS machines found in the house of a Smartmatic technician in Antipolo, Rizal after the May 10, 2010 elections.)

Christian Monsod, former Comelec Chairman: Our automation was mass-produced in one step, was not really pilot-tested satisfactorily, and was provided by a supplier who had no extensive experience in the technology and seemed to be also learning while it was being implemented. (“The 2010 Automated Elections – An Assessment,” Nov.9, 2010)

Philippine Computer Society (PCS) report on the Biliran fraud case: “There were highly questionable instances where the probability of fraud may have been perpetrated using the PCOS machines …(based on) the audit trail of the PCOS machines and the computer audit trail of the Municipal Board of Canvassers (MBOC). …There were inconsistent protective counters displayed by the PCOS machines.”

Asian Network for Free Elections (ANFREL): “Secrecy of the ballot was virtually impossible…The operation of PCOs machines encountered numerous technical glitches, ranging from minor ones that lasted for a few minutes, to others that led to suspension of polling for hours. Delay and failure of the transmission of the elections results was widely reported.”

Carter Center (U.S.): “Review of the AES's source code was limited to an on-screen presentation conducted under the auspices of COMELEC. Concerned that such a review was inadequate to identify potential problems, many watchdog groups chose to forgo participation and were unable to review the code at all….(There was) significant curtailment of the right to vote by secret ballot in the 2010 elections…(Comelec should) conduct pre-election testing in a real-world and set at an earlier date in order to ensure adequate time to correct any issues identified.

Healing Democracy: The May 10, 2010 election was no different from previous fraudulent, anomalous, and violence-ridden polls in the country. Like in previous elections, Comelec should explain for the technical glitches, transmission failures, as well as incidents of fraud and violence taking place across the country. These election

32

per

vo

ter

rule

; vo

ter

auth

enti

cati

on

to

ok

to

o l

on

g; P

CO

S

bre

akd

ow

ns

(do

cum

ente

d c

ases

)

Vo

tin

g p

erio

dsc

ien

tifi

c b

asis

. It

was

m

ore

of

a co

st-b

ased

d

ecis

ion

(at

wh

at

po

int

it w

ill

be

less

co

stly

th

an D

RE

). N

o

seri

ou

s q

ueu

ing

stu

dy

an

d s

imu

lati

on

wer

e d

on

e to

get

a p

ictu

re

of

ho

w t

he

vo

ter

lin

es

wil

l b

ehav

e o

n

elec

tio

n d

ay a

nd

ar

riv

e at

th

e o

pti

mu

m

nu

mb

er

of

PC

OS

un

its.

that

th

e si

zin

g

met

ho

d (

to d

eter

min

e n

um

ber

of

PC

OS

u

nit

s n

eed

ed)

was

u

tter

ly w

ron

g. I

t re

sult

ed i

n

“vo

lun

tary

” v

ote

r d

isen

fran

chis

emen

t d

ue

to t

he

lon

g

wai

tin

g t

ime.

Th

e in

effi

cien

t v

ote

r id

enti

ty v

erifi

cati

on

st

ep a

gg

rav

ated

th

e si

tuat

ion

fu

rth

er.

Vo

ter

auth

enti

cati

on

&

id

enti

ty v

erifi

cati

on

Th

e V

ote

r R

egis

trat

ion

S

yst

em, a

9-y

ear-

old

p

roje

ct w

hic

h h

as c

ost

at

aro

un

d P

hP

4bn

, is

stil

l in

effi

cien

t an

d

inco

mp

lete

Co

mel

ec i

s so

p

assi

on

atel

y b

iase

d

tow

ard

th

e au

tom

atio

n

of

the

vo

tin

g /

ca

nv

assi

ng

pro

cess

es

giv

ing

les

s p

rio

rity

to

th

e v

ote

r re

gis

trat

ion

sy

stem

, a p

roje

ct

wh

ich

sta

rted

as

far

Eq

ual

pri

ori

ty m

ust

be

giv

en t

o t

hes

e tw

o

core

an

d m

issi

on

-cr

itic

al f

un

ctio

ns

of

the

Co

mm

issi

on

to

re

mo

ve

the

ver

y

vis

ible

mes

sy

dy

sfu

nct

ion

we

saw

in

M

ay 2

010

bet

wee

n

6throughout the country during the May 2010 elections clearly highlighted the aws in the setup and internal security of the automated system, as well as the inadequacies in personnel training, voter's education, and contingency planning. Problems and issues encountered at the various stages of the election process, from voting and counting to canvassing and proclamation, have been recorded/documented in the reports of print and broadcast media, the hearings of the Committee on Suffrage and Electoral Reforms of the House of Representatives, the findings of the Forensic Team constituted by the Joint Congressional Canvassing Committee to examine certain PCOS machines, and the testimonies of various election stakeholders. (Post-Election Report of AES Watch: Recap and Validated of the STAR Card Assessment of the Preparations for the May 2010 Automated Elections, October 2010)

CenPEG: There was a high incidence of technical hitches, blunders, voting procedural errors, and other operational failures throughout the country during the May 10, 2010 automated elections. As The CenPEG Report reveals, these can be attributed to the lack of safeguards, s e c u r i t y m e a s u r e s , a s w e l l a s t i m e l y a n d e f f e c t i v e continuity/contingency measures (software, hardware, technologies, and other system components) that proved damaging to the accuracy, security, and reliability of election returns. Lacking these vital mechanisms, the automated election system (AES) that was harnessed for the May 10 polls was vulnerable not only to various glitches and management failures but also electronic cheating including possible pre-loading of election results. The Comelec is called upon to disclose all election documents – public information – to test and validate its claim of election “success” and debunk allegations of electronic fraud – all for the sake of public interest and voters' rights. (The CenPEG Report, December 2010)

NAMFREL: The preoccupation with new technology and speedy counting to impress the public came at the expense of greater transparency and accountability of the system… In spite of the automation of the voting process, traditional forms of electoral fraud such as vote buying; ballot capturing; use of minors in the campaign as well as in vote buying; threats, intimidation, and violence; and non-observance of secrecy and privacy during voting, remained rampant. (Terminal Report, July 10, 2010)

Joint Forensic Team (Final Report to Congress' Joint Canvassing Committee, June 10, 2010): “…The published hash code (in the Comelec website) is not the same as the extracted hash code”; “no

31

bac

k a

s A

pri

l 20

03

and

wh

ich

to

dat

e is

st

ill

a w

ork

in

pro

gre

ss

them

. If

we

sim

pli

fy

the

fun

dam

enta

l re

qu

isit

es o

f a

tru

stw

ort

hy

ele

ctio

n

we

wil

l ar

riv

e at

on

ly

thre

e co

mm

and

men

ts

(sim

ple

r an

d l

ess

nu

mb

er t

han

th

ose

fo

r g

etti

ng

to

hea

ven

):

1) f

airn

ess

(on

e v

ote

p

er q

ual

ified

vo

ter)

, 2)

pri

vac

y o

r se

cret

b

allo

ts (

vo

tes

kn

ow

n

on

ly t

o t

he

vo

ter)

, an

d

3) a

ccu

racy

(v

ote

s ar

e re

cord

ed a

s in

ten

ded

an

d c

ou

nte

d a

s re

cord

ed).

Du

e to

th

is m

essy

V

ote

rs R

egis

trat

ion

S

yst

em c

om

man

dm

ent

#1

was

mo

st l

ikel

y

vio

late

d

7In its report on “The Conduct of the Automated 2010 National and

Local Elections” released only on March 7, 2011, Comelec dubbed the May 2010 AES as credible and successful. However, it downgraded its claim of “resounding success” to ”qualified success” later. The report of CAC (June 2010) described the AES as “not a perfectly executed exercise” but, “despite the mistakes committed by Comelec and Smartmatic-TIM” it “ultimately did work.” However, since the AES has too many problems that need to be solved, it recommended to Comelec not to exercise the option to purchase the Smartmatic-TIM election system for the May 2013 elections.

The House Committee on Suffrage and Electoral Reforms (CSER) in its June 2010 report called its assessment of the AES “a mixed success”: “Automation showed no substantial advantage. On the local level, our assessment is of profound unease.” On the other hand, the JCOC on AES is mandated by law to review the automated polls within one year and recommend whether to use the same technology or a new one. But it convened finally after 2 ½ years only on Nov. 21, 2012. No assessment was ever made until Congress adjourned for the May 2013 polls.

H. What other agencies and groups were involved in the monitoring and assessment and what were their findings?

Citizens' watchdogs, organizations, and institutions that were actively involved in the monitoring and assessment included: the Automated Election System Watch (AES Watch), a broad multi-sectoral group composed of 45 organizations; the Center for People Empowerment in Governance (CenPEG); National Movement for Free Elections (NAMFREL); WeWatch (Workers Election Watch); People’ s International Observers Mission(PIOM) - National Council of Churches in the Philippines (NCCP); Kontradaya; Philippine Computer Society (PCS); and DLSU-College of Computer Studies. Others were the Consortium of Christian Organizations for Rural-Urban Development (Concord through Healing Democracy); Asian Network for Free Elections (ANFREL), Carter Mission, Global Filipino Nation, and others.

I. What major findings and conclusions (or specific highlights) were made by these groups in the monitoring and assessment of the May 2010 AES?

AES Watch: The major concerns raised by AES Watch through its STAR Card were strongly validated by the actual experience in implementing the AES. The glitches, errors, and deficiencies observed

30

Bal

lot

issu

ance

PC

OS

-dri

ven

fak

e b

allo

ts d

etec

tio

n

feat

ure

no

t av

aila

ble

; au

tom

atic

fak

e b

allo

t d

etec

tio

n m

and

ato

ry

feat

ure

did

no

t w

ork

; w

ork

-aro

un

d s

olu

tio

n

usi

ng

han

d-h

eld

UV

sc

ann

ers

was

in

effe

ctiv

e.

Sm

artm

atic

was

al

low

ed t

o u

se u

sed

p

rin

ters

fo

r p

rin

tin

g

the

bal

lots

. Th

is

cau

sed

fai

lure

in

p

laci

ng

an

eff

ecti

ve

solu

tio

n o

n t

he

bal

lots

th

at w

ill

be

det

ecte

d

by

th

e P

CO

S

auto

mat

ical

ly u

po

n

feed

ing

of

the

bal

lots

.

Th

e au

tom

atic

fak

e b

allo

t d

etec

tio

n

feat

ure

is

a M

US

T

sin

ce a

ny

hu

man

-d

riv

en d

etec

tio

n

op

tio

n w

ill

no

t b

e st

rict

ly c

arri

ed o

ut

by

th

e B

EIs

as

wh

at

hap

pen

ed. M

ost

lik

ely

fa

irn

ess

was

vio

late

d.

Act

ual

vo

tin

g

(sh

adin

g)

Vo

ter

pri

vac

y w

as

com

pro

mis

ed d

ue

to

the

len

gth

of

the

bal

lot;

oth

er v

ote

rs c

an

easi

ly s

ee o

ne'

s v

ote

s.

Th

e p

oss

ibil

ity

of

a p

riv

acy

co

mp

rom

ise

cau

sed

by

th

e lo

ng

b

allo

t d

id n

ot

even

co

me

to t

he

min

ds

of

the

Co

mel

ec

imp

lem

ente

rs.

Vo

ter

pri

vac

y o

r v

ote

se

crec

y i

s a

bas

ic r

igh

t o

f th

e v

ote

r an

d w

as

com

pro

mis

ed b

y

Co

mel

ec.

Bal

lot

sub

mis

sio

n

to P

CO

SV

ote

ver

ifica

tio

n b

y

vo

ter

pri

or

to

sub

mis

sio

n t

o P

CO

S,

a m

and

ato

ry

Co

mel

ec i

s so

wo

rrie

d

abo

ut

the

po

ssib

le l

on

g

qu

eue

that

mig

ht

form

if

th

e v

ote

r is

giv

en a

Vo

ters

did

no

t h

ave

a ch

ance

to

rev

iew

th

e v

ote

s re

cord

ed b

y t

he

PC

OS

wh

ich

is

a

8system used is really secured, accurate, and reliable or compliant with minimum requirements of the law, as claimed by the provider and Comelec.

Pollwatchers and voters alike should prepare for the worst. What happened during the February 2, 2013 mock elections involving only a few PCOS machines and staged-managed with ill prepared “voters” and actors' names in the ballots that did not simulate the real conditions on Election Day, has provided only a glimpse of the actual scenario come May 13, 2013. Aside from the usual kits, Watchers should equip themselves NOT only with skills on how to shade and how to cast the ballot into the PCOS, but more with detailed knowledge of the vulnerabilities of the PCOS machines and the environment of the clustered precincts. For cheaters, ignorance is bliss. For enlightened voters and watchers, ignorance is a crime.

F. Why is monitoring and assessment important in the implementation of the AES?

Especially because it is a new election process with its full-blast implementation in the May 10, 2010 synchronized national and local elections, the AES should be subjected to rigorous monitoring and assessment.

Monitoring and assessment are important in order to ascertain how the AES was implemented - from preparations for the May 2010 elections, how it performed on election day, as well as post-election. Assessment is also vital in establishing whether implementation is compliant with the election modernization law (legal); follows the IT standard requirements such as security, trustworthiness, accuracy, auditability, and reliability (technical); and ensures secret voting and public counting, transparency and accountability (management).

An important aspect of the assessment is to validate the AES' goal of electoral reform particularly in minimizing if not entirely eliminating fraud as well as enhancing electoral democracy.

G. Who are mandated by law to conduct the assessment/probe of the AES?

The government bodies that are mandated by law to conduct the assessment are: Comelec, Comelec Advisory Council (CAC), and the Joint Congressional Oversight Committee (JCOC) on the AES. Congress (both the House and Senate) through its concerned committees is also mandated to probe into the conduct of the AES in aid of legislation.

29

req

uir

emen

t w

as

dea

ctiv

ated

; giv

en t

he

do

ub

tfu

l ac

cura

cy o

f th

e P

CO

S, t

his

is

a v

ery

ser

iou

s is

sue.

Th

e v

ote

r d

idn

't k

no

w i

f h

is v

ote

s w

ere

corr

ectl

y r

eco

rded

.

chan

ce t

o v

erif

y

wh

eth

er h

is v

ote

s w

ere

reco

rded

co

rrec

tly

th

at

is w

hy

it

allo

wed

S

mar

tmat

ic t

o

dea

ctiv

ate

this

ver

y

imp

ort

ant

mac

hin

e fe

atu

re.

vio

lati

on

of

the

AE

S

law

. Wo

rse,

its

ab

sen

ce r

emo

ved

th

e ch

eck

th

at w

ill

ver

ify

th

e ac

cura

cy o

f th

e P

CO

S. A

vio

lati

on

of

accu

racy

– a

min

imu

m

syst

em r

equ

irem

ent.

Qu

eue

man

agem

ent

No

rea

l an

d

scie

nti

fica

lly

-bas

ed

tim

e an

d m

oti

on

stu

dy

o

f th

e en

tire

vo

ter

iden

tifi

cati

on

an

d

vo

tin

g p

roce

ss w

as

do

ne.

Th

e 5

old

p

reci

nct

s cl

ust

erin

g

dec

isio

n w

as m

ade

via

th

e se

at-o

f-th

e-p

ants

m

eth

od

(o

r w

hat

wil

l b

e ch

eap

en

ou

gh

to

h

urd

le t

he

DB

M

bu

dg

et s

cru

tin

y i

f P

CO

S i

s u

sed

).

9lawsuit filed before the Delaware chancery court revealed and validated the weaknesses and vulnerabilities of AES 2010 including the non-compliance with IT standard practices. The license issue is critical as Dominion Voting Systems can probably sue COMELEC and/or the Philippine Government for using unlicensed software for the 2013 Midterm Elections.

E. Is there anything yet to be done to salvage the wrong or correct the mistakes and ensure against fraud generated by a problematic automated election system if Comelec persists in using the provided technology?

With only two months to go (since publication of this Primer) before Election Day, the Comelec decided not to conduct anymore Mock Elections against the advice and request of the JCOC and AES Watch last February 6, 2013 for “confidence building” after the humiliating February 2 Mock Elections. With major PCOS errors still uncorrected and no source code review done by political parties and interested parties as prescribed by law, the people are left to pursue all course of actions necessary to a) assert the rule of law, b) ensure that the elections push through, c) the voters' right are not violated, d) teachers' duties and rights as BEIs are protected, and e) the integrity of the vote through automation is maintained.

Based on Chairman Brillantes' public statements, automated elections with Smartmatic as technology provider, will proceed as scheduled. As the premier election manager, the Comelec must allow stakeholders, especially political parties and election watchdogs equal access to the following MINIMUM information during election day:

1) Printed precinct audit logs2) Printed MBOC (municipal board of canvassing) audit logs3) Continuing public website (as provided for in the law) where

the canvassed votes in the national server should be posted4) Random Manual Audit immediately after elections conducted

by a competent group 5) Electronic Transmission Results

This should be contained in new Comelec Resolutions and General Instructions since only the copy of the Election Returns (ERs) are provided in RA 9369 and the outdated Omnibus Election Code as documents to be made accessible to dominant and minority political parties. All the above mentioned data are important for political parties and watchdogs to know in order to gain confidence and trust that the

28

Ex

ecu

tio

n o

f cl

ose

of

vo

tin

g p

roce

du

res;

Sh

ift

to p

ost

-vo

tin

g

mo

de;

Pri

nti

ng

of

8 E

R c

op

ies

Pri

nto

uts

of

the

ER

s (o

nly

3.5

-in

ches

wid

e an

d v

ery

, ver

y l

on

g)

are

too

sm

all

wh

en

thes

e ar

e re

qu

ired

to

b

e p

ost

ed o

n t

he

pre

cin

ct's

wal

l an

d

rem

ain

th

ere

for

48

ho

urs

aft

er p

oll

s cl

ose

d

for

the

pu

bli

c to

see

an

d r

eco

rd.

Co

mel

ec a

llo

wed

th

is

des

pit

e th

e cl

ear

inte

nt

that

th

e E

Rs

wil

l b

e p

ost

ed a

t th

e p

reci

nct

w

all

for

48 h

ou

rs a

fter

cl

osi

ng

of

po

lls.

Th

is

wil

l re

qu

ire

the

PC

OS

to

hav

e an

ex

tern

al

wid

er p

rin

ter

wh

ich

S

mar

tmar

tic/

D

om

inio

n's

pro

po

sed

so

luti

on

do

es n

ot

hav

e.

Du

e to

th

is u

nm

et

req

uir

emen

t C

om

elec

sh

ou

ld h

ave

dis

qu

alifi

ed

Sm

artm

atic

bu

t it

did

n

ot.

In

stea

d, o

ur

elec

tio

n s

yst

em w

as

mad

e to

fit

to t

he

solu

tio

n o

f S

mar

tmat

ic, n

ot

the

oth

er w

ay a

rou

nd

. T

his

les

sen

ed t

he

tran

spar

ency

of

the

elec

tio

n c

on

du

ct.

Th

is

vio

late

d f

airn

ess

and

ac

cura

cy.

Dig

ital

sig

nin

gT

he

abse

nce

of

a re

al

and

in

du

stry

-acc

epte

d

dig

ital

sig

nat

ure

fo

r

Co

mel

ec a

llo

wed

a

mac

hin

e si

gn

atu

re -

an

u

tter

ly u

nac

cep

tab

le

Its

abse

nce

co

mp

rom

ises

th

e en

tire

el

ecti

on

's i

nte

gri

ty.

10

Comelec insists that AES Watch's allegations are unmeritorious and should be dismissed outright: Comelec should not dismiss the issues as baseless. In the mock elections conducted before the HoR CSER in July, 2012 and for the public in January, 2013, the same issues and problems were observed. Perhaps some have been resolved and addressed but Comelec has failed to demonstrate the same. Many alternatives by different policy study and IT groups with reputable IT experts and lawyers, including the FIT4E or Filipino IT for Elections as well as by a few lawmakers, have been presented to Comelec, among which is a hybrid of manual voting and counting combined with electronic transmission and canvassing. But Comelec simply has turned a deaf ear and remains fixated on Smartmatic which does not even own the software to be used for the elections.

C. Based on these problems and issues, what is expected to happen in the coming May 2013 elections?

Trustworthiness, reliability, accuracy, and security among others are issues that need to be addressed. While the trusted build activity has been performed (on Jan 10, 2013), the input to the trusted build process, which is the source code of the AES, has not been reviewed by any interested political party or group. If the source code cannot be trusted, how can the output (the executable code which will be loaded in the PCOS and CCS machines) be trusted?

Further, the trusted build process covered only three sets of software. The PCOS software was not subjected to the trusted build process.

The COMELEC has been awfully silent on this issue which arises from the complaint filed by Smartmatic against Dominion Voting Systems, the owner of the PCOS technology supplied by Smartmatic-TIM.

D. If not resolved, what will be the impact of these problems

and issues on the voters and on the integrity of the May 2013 elections?

On May 23, 2012 the real owner of the AES system used in May 2010 terminated its licensing contract with Smartmatic. A subsequent lawsuit filed before the Delaware chancery court revealed and

27

the

ER

s an

d C

OC

s w

as a

ver

y, v

ery

lar

ge

secu

rity

ho

le;

any

thin

g c

an b

e d

on

e o

n t

he

resu

lts

by

an

yb

od

y w

ho

gai

ns

acce

ss t

o t

hes

e re

sult

s.

dec

isio

n f

or

no

wh

ere

in t

he

wo

rld

an

d t

he

IT i

nd

ust

ry c

an a

m

ach

ine

sig

nat

ure

q

ual

ify

as

a re

al d

igit

al

sig

nat

ure

.

Tra

nsm

issi

on

of

resu

lts

Tra

nsm

issi

on

of

the

ER

s d

id n

ot

foll

ow

th

e m

and

ato

ry

hie

rarc

hic

al s

equ

ence

an

d d

irec

tio

n t

hu

s o

pen

ing

th

em t

o

po

ssib

le f

rau

d.

Co

mel

ec a

llo

wed

un

rest

rict

ed

tran

smis

sio

n s

equ

ence

an

d d

irec

tio

n t

hu

s v

iola

tin

g w

ith

im

pu

nit

y t

his

m

and

ato

ry p

roto

col.

T

ran

smis

sio

n w

as d

on

e to

wh

ich

ever

had

an

o

pen

lin

k a

t th

e ti

me

of

tran

smis

sio

n.

Th

e co

rrec

t h

iera

rch

ical

tr

ansm

issi

on

pro

toco

l o

f fi

rst

to t

he

city

/m

un

se

rver

was

no

t fo

llo

wed

wit

h

Co

mel

ec's

ap

pro

val

; w

e d

id n

ot

kn

ow

if

the

PC

OS

wer

e tr

ansm

itti

ng

firs

t to

a

rog

ue

serv

er

som

ewh

ere

wh

ere

resu

lts

can

be

do

cto

red

o

r th

e ce

ntr

al s

erv

er

itse

lf a

lso

ser

ved

as

the

rog

ue

serv

er.

It i

s 1,

634

tim

es m

ore

11

The only WORM devices are CD or CD-R and DVD or DVD-R, both optical storage devices. Other optical storage devices are CD-R/W and DVD-R/W. The R in the nomenclature refers to RECORDABLE. The R/W in the nomenclature refers to READ/WRITE.

Authoritative sources may be found in Google.

On the “Chain of Custody:” While Comelec claims that it cannot abrograte unto

itself the duty of taking custody of ballot boxes it must be pointed out the city/municipal treasurers are under its supervisory control during the elections. Comelec failed to ensure that proper chain of custody of election paraphernalia is observed. Cases in point:

The PCOS machines in the custody of the technician who brought the said PCOS machines home

The CF cards found in a garbage dump in Cagayan de Oro City

The CF cards in the possession of certain candidates

Those who failed in the observance of proper chain of custody should have been sanctioned

Comelec's election officers, provincial election supervisors, regional election directors, and project management officers, including perhaps, the commissioners concerned , too, should have been sanctioned under the principle of command responsibility

On the wrong tally of registered voters: Indeed, the error stems from the erroneous program script. But the error greatly affected the credibility of the elections and the automated election system. This is also indicative of the poor or weak project management, analytical, and programming skills of Smartmatic's project managers, analysts, and programmers.

On the RMA sampling method: The May 2010 RMA was certainly Comelec simply does not listen to suggestions and does not engage in intelligent discussions of alternatives.

26

dif

ficu

lt t

o c

oer

ce

/co

op

t p

eop

le

man

nin

g o

ne

serv

er

site

th

an 1

,634

set

s o

f p

eop

le i

n 1

,634

sit

es

(th

e n

um

ber

of

citi

es

and

mu

nic

ipal

itie

s).

Tra

nsp

aren

cy w

as

clea

rly

vio

late

d h

ere.

Tra

nsm

issi

on

to

th

e ci

ty/

mu

nic

ipal

co

nso

lid

atio

n s

erv

er

FIR

ST

Th

e co

rrec

t h

iera

rch

ical

tr

ansm

issi

on

pro

toco

l o

f fi

rst

to t

he

city

/m

un

se

rver

was

no

t fo

llo

wed

wit

h

Co

mel

ec's

ap

pro

val

; w

e d

id n

ot

kn

ow

if

the

PC

OS

wer

e tr

ansm

itti

ng

firs

t to

a

rog

ue

serv

er

som

ewh

ere

wh

ere

resu

lts

can

be

do

cto

red

o

r th

e ce

ntr

al s

erv

er

itse

lf a

lso

ser

ved

as

the

12

of procedure in this regard. It is IT best practice that when there is a change in any part or component of the system, a Change Request document is generated. The Change Request is then routed to the appropriate parties so that the impact of such change can be analyzed. Having failed to follow and implement IT best practice, the appropriate parties within the Smartmatic organization and Comelec's Project Management Office, the team that is assigned to prepare the CF Card Configuration, were not informed of the design change.

(As a note, the CF Card Configuration includes the names of the candidates; the positions being contested; the precinct identity which includes the precinct number, the barangay, city/municipality, and province; and the number of registered voters). It should be pointed out, however, that Smartmatic claims that the problem of the CF card misconfiguration stemmed from an erroneous technology provided by Dominion (see Smartmatic complaint against Dominion Voting Systems).

The matter of the CF Card being or not being a WORM device has been a subject of debate between AESWatchers and Comelec/Smartmatic. To put this matter to rest, it is suggested that Comelec/Smartmatic demonstrate that, as they claim, the CF Card is a WORM device.

AES Watch has always held that the CF Card is not a WORM storage device. Hereunder is a technology enumeration:

CF Cards are of the same technology as USB Thumb/Flash drives, memory sticks, and SD Cards.

SD cards are used in smartphones. CF Cards are also used with Digital SLR Cameras. CF Cards, USB Thumb/Flash drives, memory sticks,

and SD cards differ only in (physical) packaging The CF Card technology base is EEPROM or

Electronically Erasable Programmable Read Only Memory (admittedly an oxymoronic term). The predecessor technology is PROM or Programmable Read Only Memory. PROMs can be written into only once. EEPROM technology is an improvement over PROM as EEPROM is rewriteable.

25

rog

ue

serv

er. I

t is

1,6

34

tim

es m

ore

dif

ficu

lt t

o

coer

ce p

eop

le m

ann

ing

o

ne

serv

er s

ite

than

1,

634

sets

of

peo

ple

in

1,

634

site

s.

To

th

e ce

ntr

al/

KB

P

serv

ers

sub

seq

uen

tly

;

Pri

nti

ng

of

22

add

itio

nal

ER

co

pie

s;

Bac

k-u

p &

sh

utd

ow

n

of

PC

OS

Th

ere

are

val

id

gro

un

ds

that

th

e C

F

card

is

no

t a

Wri

te-

On

ce-R

ead

-Man

y

(WO

RM

) st

ora

ge

med

ium

th

us

op

enin

g

the

po

ssib

ilit

y o

f ta

mp

erin

g e

ith

er t

he

resu

lts

or

the

bal

lot

con

fig

ura

tio

n a

nd

o

ther

ser

iou

s fr

aud

.

Co

mel

ec a

llo

wed

th

is

des

pit

e b

ein

g a

m

and

ato

ry r

equ

irem

ent

wh

ich

in

ten

ded

fo

r a

CD

-R o

r D

VD

-R

reco

rdin

g m

ediu

m.

Th

e S

mar

tmat

ic/

Do

min

ion

p

rop

ose

d s

olu

tio

n d

id

no

t h

ave

eith

er a

n

inte

rnal

or

exte

rnal

C

D/

DV

D d

riv

e.

We

clea

rly

saw

th

e C

F

card

mes

s th

at

hap

pen

ed 5

day

s b

efo

re e

lect

ion

day

an

d

on

ele

ctio

n d

ay i

tsel

f.

Had

Co

mel

ec a

dh

ere

to t

he

tech

nic

al

spec

ifica

tio

ns

thes

e co

uld

hav

e b

een

p

rev

ente

d.

Tra

nsp

aren

cy a

gai

n

was

a v

icti

m h

ere.

13

There are two parts in the definition of electronic signature in RA8792: (a) that an electronic signature is any electronic mark adopted by a person and (b) that the execution of an electronic signature involves the use of a procedure.

nd “Digital signature” is consistent with the 2 part of the

definition as “digital signature” technology involves the use of a procedure.

“Digital signature” technology also provides a feature of independent verification and authentication of a “digital signature”.

The “machine digital signature” implemented in the PCOS used in the 2010 Elections could not even be independently verified and authenticated.

Further, Comelec argues that in dismissing the petition filed by several parties, the SC, in its decision, stated that “the PCOS are capable of producing digitally-signed transmissions”. Indeed the PCOS machines can. But as discussed in the foregoing “machine digital signatures” are not recognized by law.

Similarly, the exchange between Justice Carpio and Atty Lazatin who are both lawyers takes root from a myopic interpretation of RA9369.

On the matter of rogue servers: It must be pointed out that the transmission should have also been done individually to the servers of: (a) the majority political party, (b) dominant minority party, (c) the accredited citizens' arm, and (d) the KBP. Instead, transmission of the ERs was made to a common server identified as the KBP server which was placed under the restrictive control of PPCRV located at Pope Pious Center at UN Avenue. This is a violation of the relevant provision of RA9369.

On the CF Card is not a WORM device: Indeed the May 3, 2010 CF Card fiasco is not at all related to the CF not being a WORM device but it is the opinion of IT security experts at AES Watch that the CF Card fiasco is a problem of misconfiguration which stemmed from the change in design of the local face of the ballot. The local face was adjusted to physically show a double-spaced listing from the original single-spaced listing. There was a failure

24

Ele

ctio

n m

ater

ials

in

ven

tory

;

Pre

cin

ct r

epo

rt

wri

tin

g;

Th

e C

hai

n-o

f-C

ust

od

y

of

sen

siti

ve

com

po

nen

ts (

bal

lots

, b

allo

t b

ox

es, C

F c

ard

s,

PC

OS

un

its,

etc

,) o

f th

e sy

stem

was

no

t se

cure

en

ou

gh

to

ass

ure

th

at i

t is

no

t b

rok

en a

ny

wh

ere

in t

he

chai

n. W

ors

e,

reco

un

ts s

tart

ed o

nly

10

mo

nth

s af

ter

elec

tio

n

day

, so

mu

ch t

ime

to

frau

du

len

tly

tam

per

th

e b

allo

ts t

o m

atch

th

e E

Rs.

Co

mel

ec s

eem

ed t

o b

e n

ot

seri

ou

s in

p

rese

rvin

g a

n

un

bro

ken

ch

ain

-of-

cust

od

y o

f se

nsi

tiv

e co

mp

on

ents

of

the

syst

em w

hen

th

is

asp

ect

defi

nes

wh

eth

er

an e

lect

ion

is

tru

stw

ort

hy

or

no

t es

pec

iall

y i

n t

he

con

tex

t o

f p

rote

st c

ases

.

Re-

pac

kin

g o

f P

CO

S

and

acc

esso

ries

;

Pre

par

atio

n f

or

RM

A i

f th

e p

reci

nct

is

sub

ject

to

RM

A;

14

Melo, preferred the use of its invented “machine digital signature” because, as he admitted in the hearings of the HoR CSER, implementation of industry recognized and accepted digital signature technology would have cost the Comelec 1billion Pesos.

Comelec's implementation of what it refers to as “machine digital signature” stems from the awed interpretation of RA9369. Atty. Jose Tolentino, interpreting RA9369 to the exclusion of all other election laws, RA8792 or the Electronic Commerce Act, and the Rules on Electronic Evidence promulgated by the Supreme Court, explained to House of Representative Committee on Suffrage and Electoral Reforms, in hearings conducted immediately following the 2010 Elections, that RA9369 did not identify who will sign on Election Returns and the Certificates of Canvass. Atty. Tolentino's explanation is quite myopic.

It should be pointed out that RA9369 amended several election laws, including among others BP881 or the Ominibus Election Code. BP881 required the BEI to sign the election returns and the BOC to sign the certificate of canvass. The provision of RA9369 where the election returns and certificates of canvass be digitally signed should have been read with BP881, among other laws. Nothing in RA9369 amended nor repealed the relevant provisions of BP881 as to who will sign the election returns and the certificates of canvass.

No Philippine Law, rule, or regulation accords legal recognition of “machine digital signature”

RA8792, which is specifically referred to in RA9369 in the manner of authenticating digital signature, accords legal recognition to electronic signatures and defines electronic signature as a signature that of a person. The principles of functional equivalence and non-discrimination between a person's handwritten signature and the same person's electronic signature is enshrined in RA8792.

While RA8792 does not make a direct reference to “digital signature”, the accepted definition “digital signature” is that it is a technology implementation of electronic signature.

23

Init

iali

zati

on

of

city

/m

un

icip

al s

erv

ers

Co

nso

lid

atio

n a

t th

e ci

ty/

mu

n s

erv

ers;

Dig

ital

sig

nin

gT

he

abse

nce

of

a re

al

and

in

du

stry

-acc

epte

d

dig

ital

sig

nat

ure

fo

r th

e E

Rs

and

CO

Cs

was

a

ver

y, v

ery

lar

ge

secu

rity

ho

le; a

ny

thin

g

can

be

do

ne

on

th

e re

sult

s b

y a

ny

bo

dy

w

ho

gai

ns

acce

ss t

o

thes

e re

sult

s.

Co

mel

ec a

llo

wed

a

mac

hin

e si

gn

atu

re -

an

u

tter

ly u

nac

cep

tab

le

dec

isio

n f

or

no

wh

ere

in t

he

wo

rld

an

d t

he

IT

ind

ust

ry c

an a

m

ach

ine

sig

nat

ure

q

ual

ify

as

a tr

ue

dig

ital

si

gn

atu

re.

Its

abse

nce

co

mp

rom

ises

th

e en

tire

ele

ctio

n's

in

teg

rity

.

Tra

nsm

issi

on

to

ce

ntr

al/

KB

P s

erv

ers

Init

iali

zati

on

of

cen

tral

se

rver

Co

nso

lid

atio

n a

t ce

ntr

al s

erv

er

15

criss-crossed the lines but there were only a maximum of 200 voters per precinct.

On the UV ink: The fact that the UV ink detection was disabled is tantamount to non-delivery of an offered / promised security feature for which Smartmatic should have been PENALIZED. The purchase of handheld UV lamp should have also been charged to Smartmatic at no cost to taxpayers' money.

On the serial numbers embedded in the bar code: The serial numbers could be used to trace the voter to the ballot. An observer with intent to identify the ballot to a voter could easily observe the issuance of the ballot and the sequence by which the voter “feeds” the ballot into the PCOS. This is a violation of vote secrecy guaranteed under the Constitution.

On voter verification: There is lack of transparency in the PCOS simply because the voter is deprived of the ability to ascertain if the machine correctly appreciated his ballot. Comelec's response (issuance of a replacement ballot in case machine appreciation of the voter's ballot is correct) indicates that the PCOS is not responsive to the requirements of our election laws.

On Comelec's comment on public verification of accuracy. The May 3, 2010 CF Card fiasco was indicative of the inaccuracy of the PCOS. Recall that the CF Cards were recalled nationwide and had to be replaced.

On the time and motion. As earlier indicated, the number of voters in a clustered precinct in 2010 was the cause of the long queue. BEIs had to attend to 5 times the number of registered voters. At what cost must we ensure that elections are conducted properly? If more PCOS machines were required, then more should have been provided.

On Digital Signature: For all its repeated claims, it is best to challenge Comelec and Smartmatic that the digital signature implemented in the AES can be independently verified and a u t h e n t i c a t e d . F u r t h e r , C o m e l e c a r g u e s o n t h e procedural/operational aspect which could have been easily addressed. Comelec, under the stewardship of Chairman Jose

22

Can

vas

sin

g a

t C

om

elec

Cen

ter

&

Join

t C

on

gre

ssio

nal

C

anv

assi

ng

Cen

ter

Th

e sh

ock

ing

ly w

ron

g

tall

y o

f th

e n

atio

nw

ide

tota

l v

ote

s (1

50+

m

illi

on

in

th

e C

om

elec

C

anv

assi

ng

ser

ver

an

d

250+

mil

lio

n i

n t

he

Join

t C

on

gre

ssio

nal

C

anv

assi

ng

ser

ver

wer

e ig

no

red

by

th

ese

two

b

od

ies;

th

is i

s an

o

bv

iou

s te

chn

ical

err

or.

Co

mel

ec e

cho

ed

Sm

artm

atic

's e

xcu

se

that

th

e w

ron

g t

ally

w

as a

res

ult

of

app

lica

tio

n e

rro

r

Th

is i

s a

BU

G o

f te

rrib

le

po

ssib

le c

on

seq

uen

ce

wh

ich

co

uld

hav

e b

een

d

etec

ted

had

a p

rop

er

ind

epen

den

t so

urc

e co

de

rev

iew

was

al

low

ed. T

he

larg

e n

um

ber

as

per

in

du

stry

p

ract

ice)

is

sup

po

sed

to

b

e th

e so

-cal

led

sy

stem

ch

ok

e p

oin

t (t

he

po

int

for

the

max

imu

m

nu

mb

er o

f b

allo

ts t

hat

ca

n b

e co

un

ted

set

as

a p

aram

eter

of

the

syst

em).

Wh

y w

ere

thes

e tw

o p

aram

eter

s se

t at

3 a

nd

5 t

imes

th

e n

um

ber

of

reg

iste

red

v

ote

rs?

Giv

en a

tu

rno

ut

of

40 m

illi

on

, wh

ere

did

th

e 11

0+ m

illi

on

an

d 2

10 m

illi

on

plu

s ex

tra

bal

lots

co

me

fro

m

16

Herewith is AES Watch's Rejoinder to the Comelec on the following problems and issues (Note: AES Watch received a copy of Comelec's 13-page reply (undated) to the AES Watch Primer first release presented to the Catholic Bishops Conference of the Philippines dated January 26, 2013):

Even if contingency plans were put in place, it is BEST practice to ascertain the cause of problems observed in an automated system and find resolutions to such problems, not just troubleshoot each problem as it comes.

For example, the issue of different timestamps on the election returns. The reason by Smartmatic was that the internal battery may have been dislodged from its receptacle during transport. Many IT experts and practitioners who have worked with PCs from the time computers started to invade the office space in the 1980s never encountered internal batteries dislodged from their receptacles. The reason given by Smartmatic is simply indicative of the low level quality of the PCOS. If their reason is true and correct, the PCOS clock should have reset itself to a common base date/time, not random date/time. Another example is the ballot does not fit the PCOS insertion slot which happened in May 2010. BEI members had to trim the the edge of the ballot to be fed successfully. Had Smartmatic determined the cause of the problem, it could have been avoided during the February 2 mock election at the UP-IS.

On the number of ballots per clustered precinct: The number of ballots is determined by the number of registered voters assigned to a polling precinct plus the number of BEI members. This could have easily been configured in the system – not a complicated thing to do. After all, Smartmatic owns the EMS and could have easily added a formula, “Number of Ballots=Number of Registered Voters+Number of BEI members,” a one line instruction that could have been added to the EMS program.

On long queues: It was not the criss-crossing of voters that resulted in long queues. No time and motion study was done by Comelec as it prepared for the 2010 elections. The long queues were a result of clustering of precincts with a maximum of registered voters at 1,000. In elections prior to 2010, voters also

21

sin

ce t

hes

e n

um

ber

s w

ere

reac

hed

by

th

e tw

o c

anv

assi

ng

cen

ters

?

Po

st E

lect

ion

Ran

do

m M

anu

al

Au

dit

(R

MA

)

Th

e R

MA

was

co

mp

lete

ly w

ron

g d

ue

to t

he

wro

ng

sam

pli

ng

m

eth

od

plu

s th

e sa

mp

les

are

no

t re

pre

sen

tati

ve

of

the

elec

tio

n c

on

test

s th

at

wer

e au

dit

ed; g

iven

su

ch, n

o d

efen

sib

le

stat

isti

cal

con

clu

sio

n

can

be

gle

aned

fro

m

the

exp

ensi

ve

effo

rt

thu

s d

efea

tin

g t

he

fun

dam

enta

l p

urp

ose

o

f th

e la

w p

rov

isio

n f

or

tran

spar

ency

.

Co

mel

ec i

s N

OT

p

roh

ibit

ed f

rom

im

pro

vin

g o

n t

he

law

p

rov

isio

n f

or

RM

A. Y

es,

it t

ried

to

im

pro

ve

the

pro

vis

ion

by

gea

rin

g f

or

5 sa

mp

les

per

dis

tric

t in

stea

d o

f o

ne

per

law

re

qu

irem

ent.

Bu

t th

e sa

mp

lin

g u

sed

was

co

mp

lete

ly w

ron

g,

hen

ce, t

he

sam

ple

s w

ere

no

t re

pre

sen

tati

ve

of

the

do

mai

ns

aud

ited

. E

rgo

no

use

ful

or

val

id

con

clu

sio

n c

an b

e m

ade

fro

m t

he

RM

A r

esu

lts.

It

en

ded

up

as

just

an

ex

pen

siv

e an

d w

aste

ful

mad

e-fo

r-m

edia

th

eatr

e o

f ze

ro a

ud

it v

alu

e.

172.1. Source Code Review2.2. Use of digital signature2.3. Voter verification2.4. Use of storage devices 2.5. Other technical provisions like accuracy 2.6. Pre-tested technology2.7. Certification and role of TEC (Technical Evaluation

Committee)2.8. Role of the CAC (Comelec Advisory Council)

3. Major implementers including the Comelec and its major citizens' arm accredited for major technical tasks like conducting random manual audit (RMA) of the system lack knowledge, competence, and appreciation of the technical complexities.

4. Safeguards (and therefore, non-compliance and violations of the laws, TOR, agreements and contract) to ensure security, accuracy and reliability of the system are conveniently set aside or disregarded and made as excuses to give way to “lower cost,” “lack of time,” and “better alternative to manual (defined as “fraudulent”) elections.”

5. Sovereignty in running the elections is compromised when the implementers rely too heavily on a foreign technology provider which does not own the operating license of the technology. (Note: the case between Smartmatic (technology provider in the Philippines vs Dominion Voting System (PCOS technology owner) remains unsettled in Delaware, USA since it was filed in September 2012.

Because of these, VOTE buying is no longer a major problem in automated elections. With the lack of major safeguards in the hardware and software components of the system, the problems in PILAHAN (long queues), BILANGAN (inaccurate counting), at BENTAHAN (not just vote buying but program system buying) are emerging as modern-day problems in election. If left unchecked, automated election forms of cheating like CF card buying, PCOS machine buying and transmission jamming plus the clustered precinct long queue-delaying tactics will prevail alongside traditional forms of cheating, wholesale fraud and violence.

20

Pro

clam

atio

nP

rocl

amat

ion

s w

ere

no

t ef

fect

ed a

s ca

refu

lly

as

they

sh

ou

ld b

e re

sult

ing

in

pro

test

s n

um

ber

ing

big

ger

th

an

in p

ast

man

ual

el

ecti

on

s.

Pro

test

sP

rote

sts

wer

e in

eff

ect

futi

le f

or

the

reco

un

ts

wer

e d

on

e 10

mo

nth

s af

ter

elec

tio

n d

ay, m

ore

th

an t

ime

eno

ug

h t

o

lose

th

e au

dit

val

ue

of

the

resu

lts.

Bo

tto

mli

ne,

th

e 20

10

AE

S d

uri

ng

its

use

on

el

ecti

on

day

an

d a

fter

b

ecam

e a

bla

ck b

ox

in

fest

ed w

ith

so

man

y

wo

rms

wh

ich

ate

its

in

teg

rity

, sec

uri

ty a

nd

tr

ansp

aren

cy.

18II. Monitoring and assessment of the AES from May 10, 2010 –

present

A. What are the critical issues on the AES?

1) Legal – Lack of compliance with provisions of law, like (1) the use of digital signatures to sign the election returns and certificates of canvass and (2) review of the AES source code by interested political parties and groups.

2) Technical – Use of CF cards is insecure. A write-once-read-mean storage medium was specified in the Request for Proposal for the automation of the 2010 elections. CF cards will be used again in 2013, in violation of COMELEC's defined requirement. CF cards can be transplanted with new data.

3) Management – Considering that the same number of PCOS machines will be deployed for the May 2013 elections, clustering of precincts will be maintained. COMELEC has to prepare for more voters per voting precinct, perhaps assign more BEI members for voter verification.

B. What have been main problems in the implementation of the Election Modernization Law or the Automated Election Law (RA 9369) since 2008?

1. Lack of transparency by the premier election manager in the country, the Commission on Elections (Comelec) in dealing with various groups of citizens' election watchdogs, Filipino IT community, research groups and other stakeholders when it comes to ensuring safeguards for security, accuracy and reliability of the chosen technology in the automated election system (AES) used for elections.

1.1. The right to know and access public information on the AES has become an exclusive domain of only a few favored or “approved” individuals and/or groups;

1.2. The Filipino IT community on the whole continues to be excluded and treated as second class to foreign provided technology and know-how.

2. Inconsistent and varied interpretations of provisions of the Poll Automation Law (RA 9369), among which are the following:

19