This is what Matt Cutts had to say:

"Hi Scott. Google does use algorithms and different techniques to block excessive automated queries and scraping, especially when a someone is hitting Google quite hard. The reason is that scraping consumes server resources. We don't want real users to be slowed down or affected just because a bot is sending bunches of automated queries to Google.

We do turn off a number of tools/bots/IP addresses that scrape us too heavily. It's a common enough phenomenon that we did a blog post on Google's Online Security Blog about the subject. In fact, I know that just a week or so ago our algorithms turned off an IP belonging to one of the entities that you mentioned in your post.

In general, I would approach the bizdev folks at Google about how to send automated queries to Google with permission. Failing that, be aware that if a tool sends too many queries to Google, we do reserve the right to disable the IP address(es) of that tool. One thing I would *not* recommend is that if a tool is blocked for bad behavior, trying to make the tool more "sneaky" (e.g. trying to make the tool look closer to a web browser). Attempts to fake out Google and pretend to be more like a web browser (after you've been blocked once already) is an example of the sort of thing that is really bad in our opinion. "

JohnMu further stated:

"This is an interesting topic. I'm not quite sure how all those links apply to our terms of service, which do not allow these kinds of automated tools.

A tool accessing other websites should try to obey the rules set forth by that website. In general, these rules are described in several ways:

1. The robots.txt covers which URLs may be accessed and which ones are disallowed. You'll notice that in our robots.txt we explicitly disallow "/search", which is what most of the ranking tools generally try to access.

2. The server result codes give more information when a URL is accessed. When our network recognizes automated queries, it may return a result code of 500 or similar.

3. The HTTP headers returned by the server can provide information through the "x-robots-tag".

4. A HTML page may provide information through a "robots" (or in our case, "googlebot") meta tag.

As far as I am aware, there are no "SEO-tools" that have permission to access our web-search results in an automated way. I am also not aware of any plans to change that in the near future."