In the OP and in your reply above, you use the terms "exact" and "case insensitive" in the context of regex matching in a way that leads me to wonder if you understand that case insensitive matching is inherently inexact. E.g.:

c:\@Work\Perl>perl -wMstrict -le
"my $rx = qr/^data_123$/i;
 print $rx;
 ;;
 for my $s (qw(data_123 DATA_123 dAtA_123  data_123x data_12 nada_123)
+) {
   if ($s =~ $rx) {
     print qq{'exact' match to '$s'};
     }
   else {
     print qq{NO match to '$s'};
     }
   }
"
(?^i:^data_123$)
'exact' match to 'data_123'
'exact' match to 'DATA_123'
'exact' match to 'dAtA_123'
NO match to 'data_123x'
NO match to 'data_12'
NO match to 'nada_123'
[download]

Update: I should have mentioned that the example above was run under Strawberry Perl version 5.14.4.1.

My bad, I reinitialized the variable as part of debugging, no wonder it would not match.

Thanks for clarifying on its use. That said, the stringify version has ?^ in front, I know that ? means 0 or more of the chars preceding it, but what does ?^ mean?

A stringized regex object is wrapped in a non-capturing grouping that preserves the regex modifiers in effect when the regex was compiled. This allows a regex object to be unaffected by the modifiers in effect in any enclosing regex 'scope' during interpolation within another regex or regex object. Consider:

c:\@Work\Perl>perl -wMstrict -le
"my $rx = qr/(?^xms:^data_123$)/;
 print $rx;
 ;;
 my $ry = qr/^data_123$/xms;
 print $ry;
 ;;
 my $rz = qr{ T $rx U $ry V }xms;
 print $rz;
"
(?^:(?^xms:^data_123$))
(?^msx:^data_123$)
(?^msx: T (?^:(?^xms:^data_123$)) U (?^msx:^data_123$) V )
[download]

See also the discussion of  "(?^aluimsx:pattern)" in the Extended Patterns section of perlre.

?^ is garbage. (?^i:...), on the other hand, means that the pattern within should be matched case-insensitively as you requested.