Internet Filtering= in Public Libraries 2007 Survey Report

Public l= ibraries=20 have also taken steps to provide safe and secure environments for

children= .=20 Through the Australian Library and Information Association (ALIA),=20 they

have=20 collaborated closely with NetAlert, the national Internet safety=20 agency.

NetAlert= =E2=80=99s=20 primary objective is to promote a safer Internet experience, particularly= =20 for

young pe= ople and=20 their families, and this objective is shared by public libraries.

Filterin= g=20 software selectively controls what content Internet users can view and=20 what

activiti= es they=20 can participate in, using a variety of automatic technologies and=20 set

parameters.

This sur= vey of=20 Australian public libraries is the third such survey1 conducted by ALIA.

It addre= sses the=20 use of filtering in public Internet access terminals. The survey

instrume= nt was=20 developed to provide feedback to the Department of Communication,

IT and t= he Arts=20 (DCITA) and the general public in relation to the Australian

Governme= nt=E2=80=99s=20 Protecting Australian Families Online (PAFO) initiative announced=20 in

June=20 2006.

Responde= nts=20 answered questions relating to their libraries=E2=80=99 provision of publi= c

Internet= =20 services, their experiences of filtering software, and their attitudes=20 towards

the use = of=20 filtering in public libraries. A more extensive survey will be conducted=20 in

late 200= 7.=20

Ke= y=20 findings

=E2=80= =A2 There were 104 respondents, out of a potential total of 548 p= ublic=20 library

services= in=20 Australia.

=E2=80= =A2 39% of respondent libraries currently use filtering software = on some=20 or all of

their li= braries=E2=80=99=20 public Internet access terminals (compared to 31% in 2005 and

18% in=20 2002).

=E2=80= =A2 Respondents had largely negative views of Internet filtering.= They=20 frequently

characte= rised it=20 as a form of censorship, a limit on intellectual freedom, and

contrary= to the=20 ethics of librarianship.

=E2=80= =A2 Respondents frequently reported experiences of, and concerns= =20 about,

unreliab= ility=20 and inaccuracy of filtering software.

=E2=80= =A2 Respondents favoured educating the public for safe Internet a= ccess,=20 informing

users of= the=20 conditions of use, requiring their consent to these conditions, and=

less for= mal=20 methods of monitoring behaviour in libraries. These practices were<= /DIV>

preferre= d to=20 implementing the perceived =E2=80=98surveillance=E2=80=99 technology of=20 filtering.

1 Australian Library and Information Association s= ummary=20 report on Survey of Internet acc= ess in=20 public libraries,

2002 = ,=20 http://alia.org.au/advocacy/Internet.access/summary.report= .2002.html=20 and Internet access in public

libra= ries=20 survey 2005 summary report http://alia.org.au/advocacy/Internet.access/survey.results= .2005.pdf

Page 4

Australi= an=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 2

=E2=80= =A2 40% of libraries surveyed received Internet access through a = wider=20 council

network,= with=20 85% providing networked PCs. This has implications for the

design=20 requirements of filtering software for public=20 libraries.

=E2=80= =A2 70% of those operating filtering software deployed it a serve= r=20 level.

=E2=80= =A2 If filtering software had been or was to be installed, librar= ians=E2=80=99=20 primary

requirem= ent was=20 for easy, localised control. Respondents wanted the

followin= g=20 software features (a) the ability to choose content types to block<= /DIV>

content,= (b) the=20 ability to prevent or allow access to email, games or chat, and

Se= lected=20 responses to Internet filtering software in public=20 libraries

=E2=80= =98In the last=20 three years we have had only five instances of unacceptable content=20 being

accessed= on our=20 network =E2=80=93 the introduction of filtering software is obtrusive and= =20 unwarranted

and will= have a=20 negative impact on legitimate research.=E2=80=99

=E2=80= =98An overwhelming=20 majority of our users are responsible users of the Internet, who abide=20 by

the libr= ary's=20 own Internet policies which have been developed with the principles of=20 freedom

of acces= s to=20 information in mind. The education of parents in their responsibility to=20 monitor

their ow= n=20 children's use of the Internet would be a more constructive and beneficial= use=20 of

resources.=E2= =80=99

Conclus= ions

While th= e survey=20 provided some information about the technical aspects of

implemen= ting=20 Internet filtering software in public libraries, respondents amply<= /DIV>

expresse= d their=20 opposition to filtering per se, based on their experience,=20 professional

ethics a= nd sense=20 of the purpose of libraries.

Although= public=20 Internet filtering is in use in a significant number of public=20 libraries,

most lib= rarians=20 have concerns about the implications of any restriction on the free=

flow of= =20 information, and see filtering as problematic. Current filtering technolog= y=20 does

not meet= the=20 requirements of public libraries in terms of accuracy and reliability.=20 The

experien= ce of=20 most library professionals is that existing norms, sanctions and

informal= =20 monitoring, combined with the continued education of users about safe=20 and

appropri= ate=20 Internet use, are sufficient to combat the actual level of threat posed=20 by

inapprop= riate=20 Internet use in public libraries.

ALIA rec= ommends=20 that to provide a professionally acceptable and useful service,

Internet= =20 filtering software would need to be developed in close cooperation=20 with

public=20 librarians, with features maximising local control and adaptability to net= works.=20

ALIA wil= l=20 continue to work with NetAlert to further the education of=20 parents/carers,

children= and the=20 community regarding Internet safety.

Page 5

Australi= an=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 3

Repor= t=20

Introdu= ction

Public l= ibraries=20 are crucial providers of public Internet access through 548 services=20 in

more tha= n 1754=20 locations2. The daily lives of Austr= alian=20 citizens are increasingly

located = in an=20 interwoven environment of electronically mediated and traditional

interact= ions.=20 The public enjoy, and often rely on, free Internet access in their<= /DIV>

librarie= s.=20

In the l= ast two=20 years librarians have witnessed the emergence and explosive growth<= /DIV>

of parti= cipatory=20 Internet technologies referred to as =E2=80=98Web 2.0=E2=80=99 and, in lib= raries,

=E2=80= =98Library=20 2.0=E2=80=993. The Internet is now i= ncreasingly a=20 space for bottom-up content

creation= ,=20 complex personal interactions, and the potential enhancement of a

democrat= ic=20 =E2=80=98public sphere=E2=80=99. Public libraries=E2=80=99 provision of In= ternet access in=20 this

emerging= =20 environment is a vital service, by which they can facilitate=20 patrons=E2=80=99

connecti= on with=20 e-government, expanded contributions to community life, and the economy.

However,= the=20 Internet is also seen as a vehicle for potential threats to library=20 users,

children= in=20 particular. Public concerns about unrestricted access to the=20 Internet

include = the=20 following uses:

=E2=80= =A2 to access violent/terrorist/hate material;

=E2=80= =A2 to target vulnerable people for potential manipulation, explo= itation=20 and

violence= via=20 instant messaging (IM) or =E2=80=98chat=E2=80=99;

=E2=80= =A2 to commit financial fraud, identity theft and other crimes;=20 and

=E2=80= =A2 to access legal and illegal=20 pornography.

Because= =20 libraries provide Internet access in a public space, funded by=20 public

monies, = these=20 concerns raise the question of the role of government in ensuring

public s= afety.=20

A NetAle= rt=20 survey of children=E2=80=99s use of the Internet in April 2005 demonstrate= d that=20 the

majority= of=20 Australian=E2=80=99s aged eight to thirteen accessed the Internet either a= t home=20 or

at schoo= l.4

Since 19= 99, the=20 Australian Government has implemented several policies to promote

and faci= litate=20 safer and more secure use of the Internet, including NetAlert and StaySmartOnli= ne5.

Filterin= g=20 software selectively controls what content Internet users can view and=20 what

activiti= es they=20 can participate in, using a variety of automatic technologies and=20 set

paramete= rs.=20

Australi= an=20 Bureau of Statistics 8561.0, Public Libraries, Australia,=20 2003-04.

3 See http://en.wikipedia.org/wiki/Web_2.0 and http://en.wikipedia.org/wiki/Library_2.0 for more information.

4 Kids online@home http://www.netalert.net.au/02010-kidsonline@home---Interne= t-use-in-Australia-homes---

April-2005.pd= f

5 http://www.netalert.net.au, http://www.staysmartonline.gov.au/.

Page 6

Australi= an=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 4

ALIA has= a clear=20 policy position on Internet Filtering6. It=20 supports the basic right of all

library = and=20 information services=E2=80=99 users to unhindered access to information

regardle= ss of=20 format. This position is based on the principle that freedom can be=

protecte= d in a=20 democratic society only if its citizens have unrestricted access to=

informat= ion and=20 ideas. Thus ALIA urges that access to electronic information

resource= s in=20 libraries should not be restricted, except as required by law, and=20 this

basic ri= ght=20 should not be eroded in the development of regulatory measures for<= /DIV>

online=20 information.

While AL= IA=20 supports the =E2=80=98Protecting Australian Families Online=E2=80=99 (PAFO= ) initiative=20 for

home use= , it=20 does not recommend the use of Internet filtering technology in=20 public

librarie= s. ALIA=20 has worked with organisations such as NetAlert in developing well-<= /DIV>

publicis= ed=20 educational and information materials for libraries and Internet users=20 (for

example,= =20 NetAlert is a sponsor of ALIA=E2=80=99s Library and Information Week 2007)= .=20

Australi= an=20 public libraries have a range of strategies to ensure that public=20 library

environm= ents are=20 safe, especially for children. These include, as well as use of

commerci= al=20 filters in some cases, clear Internet use policies (95% of=20 libraries

surveyed= ), user=20 behaviour policies, supervision by staff, parental involvement and<= /DIV>

parent c= onsent=20 to approve Internet access rights (71% of libraries surveyed), and=20 the

availabi= lity of=20 information and training (85% of libraries surveyed)7.

As infor= mation=20 professionals, public librarians are well positioned to observe

informat= ion=20 seeking and knowledge creating behaviours, and the norms and=20 culture

of their= =20 community of users.

Ab= out the=20 survey

This sur= vey was=20 carried out via an online questionnaire (Appendix) that was

designed= by=20 staff from ALIA with the intention of providing feedback to the=20 general

public a= nd=20 DCITA. It was promoted to Australian public librarians through the= =20 ALIA

website,= ALIA=20 e-lists, the ALIA Online Content Regulation Group and the ALIA=20 Public

Librarie= s=20 Reference Group. Responses were gathered between 12 April 2007 and=20 11

May=20 2007.

The surv= ey forms=20 part of an ongoing body of research into public libraries and

Internet= use,=20 following on from ALIA=E2=80=99s Internet Access in Public Libraries 2002=20 and

2005=20 surveys8. This survey specifically a= ddresses=20 the topical question of Internet

filterin= g and=20 the technical aspects of its implementation, so statistical=20 comparisons

with pre= vious=20 surveys are limited. The IAPL 2005 survey contained 60 questions=20 and

this sur= vey only=20 included 20 questions. However some comparisons may be made,

since th= e=20 IAPL 2005 survey did gather relevant responses=20 about:

=E2=80= =A2 complaints received by librarians about Internet=20 content;

=E2=80= =A2 extent of filtering software use in public libraries;=20 and

=E2=80= =A2 complaints by patrons about the use of=20 filters.

6 http://alia.org.au/policies/content.regulation.html=

7 The report of the Internet Access in Public Libraries <= FONT=20 style=3D"FONT-SIZE: 10px">2005 survey is available=20 at
http://alia.o= rg.au/advocacy/Internet.access/survey.results.2005.pdf=20 .

8 The report of the Internet Access in Public Libraries <= FONT=20 style=3D"FONT-SIZE: 10px">2005 survey is available=20 at
http://alia.o= rg.au/advocacy/Internet.access/survey.results.2005.pdf=20 ; a summary of the 2002 survey is=20 available

at http://alia.org.au/advocacy/Internet.access/summary.report= .2002.html=20 .

Page 7

Australi= an=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 5

The surv= ey=20 design elicited qualitative and quantitative information, providing=

responde= nts with=20 opportunities to express fully developed opinions. The researchers<= /DIV>
obtained= data=20 about respondents=E2=80=99 experiences of and thoughts about Internet

filterin= g in=20 public libraries. 108 questionnaires were entered, yielding a solid=20 sample

(104 usa= ble=20 responses, without duplicates and spam). This sample is broadly
represen= tative=20 of the range of public library services throughout Australia,=20 including

state an= d local=20 libraries and large metropolitan and small regional libraries. Some=

responde= nts=20 wrote about their individual branch workplaces, while others=20 provided

data abo= ut=20 multi-branch library services.

Surve= y=20 Respondents by State/Territory
States/ter= ritories

Numbe= r=20 of
respondent= s
%

ACT

1=

0.9

NSW

22
21.1
NT

2=

1.9

Qld

20
19.2
SA

24
23.1
Tas

1=

0.9

Vic

20
19.2
WA

14
13.4 Total<= /NOBR>

104

100

Libra= ry=20 service location/type
Location/t= ype

Numbe= r=20 %
Metropolitan<= /NOBR>

55=20 52.9

Greater= =20 metropolitan

24=20 23.1
Regional/rura= l

20=20 19.2

No=20 data

5=20 4.8

Total= =20

104=20 100

Inter= net=20 access points by State/Territory
State/terr= itory

Numbe= r of=20 terminals

ACT

77

NSW

540

NT

9=

Qld

968

SA

344

Tas

158

Vic

760

WA

212
Total<= /NOBR>

3068= =20

Page 8

Australi= an=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 6

Surve= y=20 Results

Three do= minant=20 themes can be identified in the responses gathered.

1. A spe= ctrum of=20 largely negative views on Internet filtering, from professional
concern = to=20 outright rejection, seeing it as a form of censorship and a limit=20 on

intellec= tual=20 freedom

2. A wid= e range=20 of experiences of, and concerns about, the unreliability and

inaccura= cy of=20 filtering software

3. A tre= nd to=20 increase in filtering in public libraries accompanied by scepticism=

about th= is=20 approach. Of those surveyed, 72% of respondents had received

complain= ts or=20 expressed concerns about the use of filtering software, 11%

with few= or no=20 complaints and 17% provided no response.

On the t= echnical=20 aspects of deploying Internet filtering, the survey respondents
provided= =20 inconsistent information. This suggests that the questions and/or
terminol= ogy may=20 not have been understood in some cases, or the respondents did

not have= the=20 most accurate information on some topics, or were not the people=20 who

dealt wi= th these=20 issues.

Cu= rrent=20 use of filtering software

Approxim= ately=20 39% of respondents indicated that their library currently used=20 Internet

filterin= g=20 software on some or all of its public Internet access terminals.=20 Previous

surveys= =20 indicated levels of 31% (2005) and 18% (2002).
<= FONT=20 size=3D3 face=3DTimes>
Curre= nt use=20 of internet filtering software

No

60%

Yes on=20 all
machines

33%

Yes on=20 some
machines

6%

No data= =20 entered
1%

Page 9

Austral= ian=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 7

Respo= nses to=20 the Issue of Filtering

Re= asons=20 for not installing content filtering=20 software

Many lib= rarians=20 gave extensive comments and had strong opinions against installing<= /DIV>
filterin= g=20 software on their public Internet access terminals.

Two key = themes=20 stood out:

=E2=80= =A2 The imperative for an unrestricted flow of information and of= =20 intellectual
=
freedom = in=20 public libraries, and librarians=E2=80=99 opposition to censorship=20 (mentioned

25=20 times).

=E2=80= =A2 The inadequacy and unreliability of filtering, in that filter= s often=20 prevent access

to legi= timate=20 content and allow undesirable content through (mentioned 27

times).= =20

This se= cond=20 theme emerged in the 2005 survey, which reported that, with the
increas= e in use=20 of filters between 2002 (18%) and 2005 (31%), there had been a

corresp= onding=20 increase in complaints about the use of filters.9

Several= other=20 themes were evident in public librarians=E2=80=99=20 responses:

=E2=80= =A2 The library already has adequate supervision and sanctions in= place,=20 and has few
<= SPAN=20 style=3D"FONT-FAMILY: Times; FONT-SIZE: 14px">
problem= s=20 without filtering. This is in accordance with the findings of the=20 Internet

Acce= ss in=20 Public Libraries 2005 survey, which found that 71% of respondents=20 had

receive= d five=20 or fewer complaints about Internet content in the 12 months

precedi= ng the=20 survey, and 50% had received no complaints at all. The 2005

survey = also=20 found that 80% of respondents reported that they monitor Internet
use, mo= stly=20 visually.10
=E2=80= =A2 Users=E2=80=99 access to appropriate content is/should be a m= atter of=20 individual and

parenta= l=20 responsibility

=E2=80= =A2 The importance of educating for safe and responsible Internet= use,=20 rather than
<= SPAN=20 style=3D"FONT-FAMILY: Times; FONT-SIZE: 14px">
impleme= nting=20 the perceived =E2=80=98surveillance=E2=80=99 technology of=20 filtering

=E2=80= =A2 Trusting the effectiveness of filtering software leads to a = =E2=80=98false=20 sense of security=E2=80=99

about t= he=20 absence of inappropriate content

=E2=80= =A2 Users are already required to accept terms and conditions to = use=20 public Internet
terminals
=E2=80= =A2 Parents are already required to provide consent in order for = children=20 to use public

Interne= t=20 terminals

=E2=80= =A2 There is concern about the staff time and funds needed to ins= tall,=20 maintain and
=
monitor= =20 filtering software.
=
9 http://alia.org.au/advocacy/Internet.access/survey.results= .2005.pdf, page. 7

10http://alia.org.au/advocacy/Internet.access/survey.results= .2005.pdf, page 6-7

Page 10
<= FONT size=3D3=20 face=3DTimes>
Austral= ian=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 8

S= elected=20 responses to key themes =E2=80=93 reasons for not installing=20 filtering

=E2=80= =98Users are=20 entitled to have freedom to access information and they must agree to abid= e=20 by

the lib= rary=20 Internet use policy. We believe that Internet filtering software is not=20 effective and

restric= ts some=20 sites which may be of legitimate use to some people.=E2=80=99

=E2=80= =98Materials=20 should not be excluded because of the origin, background, or views of=20 those

contrib= uting to=20 their creation. Libraries should provide materials and information=20 presenting

all poi= nts of=20 view on current and historical issues. Materials should not be proscribed= =20 or

removed= because=20 of partisans or doctrinal disapproval. Libraries should challenge
censors= hip in=20 the fulfilment of their responsibility to provide information and=20 enlightenment.=E2=80=99

=E2=80= =98Filtering is=20 censorship. Library X is opposed to censorship and supports unhindered=20 access

to reso= urces=20 with restrictions imposed only as required by legislation =E2=80=A6 Legiti= mate &=20 useful

informa= tion is=20 blocked by filters. ALA [American Library Association] studies indicate=20 filters

block a= s much=20 as 1 in 5 sites containing legal, useful=20 information.=E2=80=99
http://www.a= la.org/ala/oif/ifissues/filtersfiltering.htm

=E2=80= =98[I am]=20 philosophically opposed to censorship of information - as all librarians s= hould=20 be.=E2=80=99

=E2=80= =98[I believe]=20 that patrons should have free access to information and that they are=20 generally

intelli= gent=20 enough to make their own decisions about what they need to access. [I]=20 also

believe= that=20 parents are responsible for monitoring what their children access via=20 the
Internet.=E2= =80=99

=E2=80= =98Filtering is=20 very subjective - what may be inappropriate to some is not offensive to=20 others.

How can= =20 boundaries that are acceptable to all be set?=E2=80=99

=E2=80= =98All software=20 we have investigated in the past has been too restrictive, depriving adult= s=20 of

useful= =20 information.=E2=80=99

=E2=80= =98Filtering=20 limits access to =E2=80=9Clegitimate=E2=80=9D information. Other means of = control/supervision in=20 the

Library= can=20 prevent inappropriate Internet use.=E2=80=99

=E2=80= =98Filtering=20 systems can never be perfect. [The] main reason for not using them is beca= use=20 they

can inh= ibit=20 legitimate research. [Filtering] is a form of=20 censorship.=E2=80=99

S= elected=20 responses to other themes =E2=80=93 reasons for not installing=20 filtering

=E2=80= =98Parents are=20 responsible for their children. We should be educating children about=20 the

dangers= of the=20 Internet and the possibility of encountering inappropriate=20 material.=E2=80=99

=E2=80= =98Our PCs are=20 located in an area that is fully visible to staff which gives adequate=20 security.=E2=80=99

=E2=80= =98We have taken=20 steps to reduce the risks of inappropriate activity and communicated=20 the

accepta= ble use=20 of the Internet-enabled PCs.=E2=80=99

=E2=80= =98We clearly=20 advise that we do not use filtering software and underage users need=20 parental

consent= to use=20 the Internet. We believe that filtering does restrict access to=20 legitimate

informa= tion.=20 However, we monitor sites accessed.=E2=80=99

=E2=80= =98Adults can=20 choose themselves and use the Internet according to an agreement they=20 must

acknowl= edge=20 before using the Internet. Children under 18 are not allowed on the=20 Internet

unless = this=20 agreement is signed by their parents/legal guardian.=E2=80=99

=E2=80= =98Filtering=20 software will limit legitimate uses of the Internet and provide a false se= nse=20 of

securit= y to=20 staff and parents as they can not absolutely guarantee effective filtering= =20 of

offensi= ve or=20 inappropriate material.=E2=80=99

Page 11
<= FONT size=3D3=20 face=3DTimes>
Austral= ian=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 9

=E2=80= =98Most of the=20 issues that the government are reacting to have been from a noisy minority= =20 and

from th= e media=20 looking for beat up stories =E2=80=A6 Perhaps if the government had talked= to=20 Public

Librari= es=20 Australia (www.pla.org.au) and ALIA prior to making the PAFO announcement= =20 it

could h= ave made=20 a more informed and less knee jerk reaction decision that would=20 have

benefit= ed the=20 whole Australian community.=E2=80=99
Concer= ns/issues=20 about Internet content filtering

72% of= =20 respondents had concerns or issues about Internet content filtering,=20 while

11% exp= ressed=20 none (a further 17% gave no response).

Respond= ents=20 articulated very similar concerns to the key themes detailed above.=

Many li= brarians=20 raised concerns about filters acting to censor and restrict the=20 free

flow of= =20 information. Several regarded the filtering of public library=20 information

resourc= es as=20 contrary to their professional ethos. They regularly reiterated=20 their

unease = about=20 filtering software=E2=80=99s accuracy and reliability, and the consequent= =20 limiting

of patr= ons=E2=80=99=20 access to legitimate information.

S= elected=20 responses =E2=80=93 concerns/issues about=20 filtering

=E2=80= =98I have grave=20 concerns about filtering and the way it limits access to information. This= =20 Library

does no= t=20 support censorship which is what filtering amounts to.=E2=80=99
=E2=80= =98Teaching users=20 to navigate the Internet intelligently and educating them on what to be=20 aware

of is o= f more=20 value than taking a patriarchal approach and making value judgements on=20 their

behalf = on what=20 is offensive =E2=80=A6 As web 2.0 pervades further and social networking=20 applications

increas= e, more=20 everyday people are creating everyday content which reflects our=20 language,

which w= hether=20 we like it or not contains words that some people may find offensive.=20 Does

that im= ply that=20 all the accompanying information should be censored?=E2=80=99

=E2=80= =98[I do] not=20 [have any concerns or issues about Internet content filtering] if the soft= ware=20 is

reliabl= e and=20 effective, and adjustable to local needs=E2=80=99

=E2=80= =98[Filtering=20 software is unable] to incorporate or allow for the complexity of=20 ever-changing

languag= e and=20 cultural contexts =E2=80=A6 [filtering software] blocks legitimate sites t= hat contain=20 useful

informa= tion=20 (typical examples being information on breast cancer or AIDS) =E2=80=A6=20 [filtering

softwar= e has a]=20 negative impact on social networking technologies.=E2=80=99

=E2=80= =98In the last=20 three years we have had only five instances of unacceptable content being= =20 access

on our = network=20 =E2=80=93 the introduction of filtering software is obtrusive and unwarran= ted and=20 will

have a = negative=20 impact on legitimate research.=E2=80=99

=E2=80= =98We have=20 adopted a minimal approach to filtering, preferring instead to approach=20 customers

and sug= gest=20 that the site they are accessing is inappropriate for a public environment= .=E2=80=99=20

=E2=80= =98[My] only=20 [concern/issue is] that the control over what is filtered stays at our lev= el so=20 that we

can tai= lor it=20 to our needs. Blanket approaches rarely work, nor do government=20 mandated

blackli= sts.=E2=80=99=20

=E2=80= =98[I am] very=20 much [concerned]. Council has applied filtering on the staff networ= k and=20 it has

caused = major=20 problems for customer service in the libraries. We are constantly having=20 access

blocked= to=20 legitimate sites and having to have it addressed.=E2=80=99

=E2=80= =98Filtering is=20 not the answer to the problems we encounter at the Library; education=20 and

parents= taking=20 the responsibility for ensuring their children are aware of the dangers of= =20 the

Interne= t must=20 remain a priority. Library X actively draws the attention to parents of=20 the

"danger= s" of=20 the Internet by providing them with information, including the NetAlert=20 brochure

before = allowing=20 children access to the Internet.=E2=80=99

Page 12
<= FONT size=3D3=20 face=3DTimes>
Austral= ian=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 10

=E2=80= =98[T]here is no=20 need to treat people as if they are infantile and need to be protected fro= m=20 the

"danger= s" of=20 the Internet. Put the money that would have been wasted on this project=20 into

educati= ng=20 people on how to appropriately use the wonderful resources that is the=20 Internet.

Librari= ans have=20 a key role in this education and, as a trusted profession, anything we hav= e=20 to

say wou= ld be=20 accepted by the populace=E2=80=A6 Educate not regulate!=E2=80=99
=E2=80= =98Experience=20 indicates that content filtering is not foolproof and can provide a false = sense=20 of
security.=E2= =80=99

P= roblems=20 with previous use of filtering

Six res= pondents=20 had used filtering software in the past and then disabled it. The=20 four

who gav= e=20 reasons each referred to problems of filters=E2=80=99 inaccuracy and=20 unreliability.

S= elected=20 responses =E2=80=93 problems with filters

=E2=80= =98[The] filter=20 was based on blocking keywords. This created problems for some=20 legitimate

researc= h but=20 was also ineffective in blocking offensive sites. We experienced a lot=20 of

frustra= tion by=20 our users when they were locked out for no apparent reason=20 including

governm= ent,=20 employment agencies, banking sites =E2=80=A6. Customers lost confidence in= =20 our
service.=E2= =80=99

=E2=80= =98[The filter=20 software was] too restrictive. Health information comes in strange places = at=20 times.

Legitim= ate=20 searches [were] blocked.=E2=80=99

I= ssues=20 encountered in current operation of a filtered Internet servic= e

This qu= estion=20 elicited similar information, from librarians who had not disabled<= /DIV>
content= =20 filters. Of respondents currently operating filtering on their publ= ic=20 Internet

termina= ls, 80%=20 reported that their filtering software prevented patrons from

accessi= ng=20 legitimate content, and/or allowed through undesirable content. 14%=

reporte= d few or=20 no issues.

Sele= cted=20 responses =E2=80=93 issues with current operation of=20 filtering

=E2=80= =98Sometimes the=20 software has incorrectly classified a URL, also sites are increasingly=20 more

complex= &=20 we may only want to block the email function but not the rest of the=20 site.=E2=80=99

=E2=80= =98No complaints.=20 We filter to a very minimal level for content.=E2=80=99

=E2=80= =98Customers=20 unable to access sites they feel should be generally available.----Downtim= e=20 due

to fail= ure of=20 filtering service or database.=E2=80=99

=E2=80= =98The occasional=20 false positives and on the very rare occasion inappropriate content not=20 being

picked= =20 up=E2=80=99

=E2=80= =98Some very=20 commonly used sites are unavailable to library users. Many customers=20 have

complai= ned at=20 the difficulty in accessing certain sites.=E2=80=99

=E2=80= =98Blocks a=20 number of legitimate sites and is also not foolproof in blocking pornograp= hy=E2=80=99=20

Page 13
<= FONT size=3D3=20 face=3DTimes>
Austral= ian=20 Library and Information Association Internet Filtering in Public Libraries= =20 Survey 2007 Report | 11

Tech= nical=20 Aspects of Filtering

M= ode of=20 Internet provision

The mos= t common=20 mode of Internet provision was through a wider local council

service= (38%),=20 followed by the use of a stand-alone library network=20 (29%).

Mode= of=20 internet provision

0

5

10

15

20

25

30

35

40

45

Wider=20 local

council= =20 service
Stand-alone<= /NOBR>

library= =20 network

Other/n= o=20 data
entered

Connect= ion=20 for

each=20 terminal

Wider=20 branch

library= =20 service
N