MIMEsweeper maker loses UK High Court patent fight over 15-year-old bulletin board post

Does a review by an 'imagineer' in 2003 invalidate a patent filing? Take a guess

A commercial rival of email virus-scanning software firm Glasswall has lost its High Court attempt to use a bulletin board post from 2003 written by a former MessageLabs "imagineer" to have a patent declared invalid.

The full judgement, which dismissed the application for the revocation of the patent, was handed down on Friday at the High Court of Justice in London.

Clearswift, purveyor of the MIMEsweeper software, had sued Glasswall alleging that one of its ideas patented in 2013 was not, in fact, patentable, on the grounds that it lacked an "inventive step".

The case boiled down to Clearswift claiming that the idea at the heart of a Glasswall-owned patent titled "resisting the spread of unwanted code and data" (European Patent EP 1 891 571 B1) was not novel enough to be patented because it had already been publicly described. If Clearswift was right, this would allow it to copy the patent's function without getting stung for intellectual property infringement.

As evidence for this, Clearswift cited a previous 2005 US patent* describing email scanning as well as what deputy High Court judge David Stone described in his judgment as "a series of posts on an Internet bulletin board" under the subject line 'Avecho Glasswall Anti virus technolog?' [sic]" which can be found here.

Email content scanners normally dismantle, parse and regenerate emails before dropping them into recipients' mailboxes, occasionally deleting attachments or body text if something malicious is detected by a threat filter, as well as inserting the boilerplate "This email has been scanned by Product X" wording. The patent disputed by Clearswift described how MIME emails can be dismantled and parsed by a virus scanner, including attachments to those emails.

Dismissing Clearswift's argument that the 2005 US patent covered the same thing (parsing and regeneration of attachments) as Glasswall's patent, the judge said: "I do not accept that it would have been obvious to the skilled addressee to apply a bypass whitelist only to files which had otherwise failed the parsing/regeneration process."

As for the bulletin board thread from 2003, the key seventh post referred to by Clearswift was written by Alexander Shipp, a one-time anti-malware "imagineer" (PDF) for MessageLabs, who these days is chief techie of Equine Register Ltd. The judge ruled that the thread "is a mix of third party comments based on using the software, and the software producer's advertising about what the software can do. It was agreed that the discussion would be of particular interest to the skilled person, including because of the claims made for the software, and the reputation of the people who wrote the posts."

Clearswift argued that Shipp's post reviewing the email-scanning software in 2003 more or less described the same functions as Glasswall's patent – pre-dating their filing by 12 years and therefore making it invalid. Unfortunately for Clearswift, Judge Stone disagreed, ruling that the post "does not disclose a number of teachings of the Patent", including its essential features of removing unwanted code and regenerating the email, or "parsing to content level", though it did infer that a threat filter similar to that used by Glasswall's IP had been in use at the time.

The full judgment is available on the BAILII website. Judge Stone does not appear to have been impressed by the "628 paragraphs" of expert evidence produced for Clearswift by Shipp, or the "40 exhibits, only one of which I was taken to during the trial", waspishly commenting: "It was probably not necessary for much of it to be before the Court." ®

* United States patent application 2005/0081057A1, titled “Method and system for preventing exploiting an email message”, published on 14 April 2005.




Biting the hand that feeds IT © 1998–2019