Writing a script to act as a mozilla browser for screen scraping

I have a script which scrapes some files from a website. The site needs a username and password to login. The script ran fine for a while, but now - even though the site recognizes me on the browser when I log in through the cookie - it does not recognize  the script as being authorized to view that page.

Here is a excerpt from the script:

my $cookie_jar = HTTP::Cookies::Netscape->new(
       file => "C:\\Documents and Settings\\mcorsi.SPEARREPORT\\Application Data\\Mozilla\\Profiles\\default\\wx9k36hh.slt\\cookies.txt",
   );
   my $browser = LWP::UserAgent->new;
   $browser->cookie_jar( $cookie_jar );
my $response = $browser->get("http://www.somewebsite.com");


The response from the script is that I need to be a registered user to view that page even though when I go to the exact same URL with the  mozilla browser I receive the correct page displayed.

Any ideas?