QUOTE cite="http://dev.w3.org/html5/spec/editing.html#the-accesskey-attribute"
When the user presses the key combination corresponding to the assigned 
access key for an element, if the element defines a command, and the 
command's Hidden State facet is false (visible), and the command's 
Disabled State facet is also false (enabled), then the user agent must 
trigger the Action of the command.
UNQUOTE

note: the term "defines a command" is defined at:
http://dev.w3.org/html5/spec/commands.html#concept-command

this should be stated less obliquely:

When the user presses the key combination corresponding to the assigned 
access key for an element, if the element defines a command

    1. check if the command's Hidden state facet is false (visible)

    2. check if the command's Disabled state facet is false (enabled)

    3. if the command's Hidden and Disabled state facets are false, 
       trigger the Action of the command.


ADDITIONAL CONSIDERATION:

The above prose should be followed DIRECTLY by a note denoting that 
accesskey activation response MUST be controlled, storable and shareable 
by the user through a user agent setting. If an author defines a 
specific action to a command, users MUST be able to override the author's
suggested action, in observance of a persistent user setting. This 
allows the user to use a "safety" to avoid firing an event defined for
an element when the accesskey defined for that element is invoked, so
that a user may make an informed decision as to whether or not to 
activate the element or to move focus to that element in order to 
inspect it for pre-set triggers.