Software patents "lost in lawyer land"

Or at least so says the EPO, according to Lucy Sherriff at The Register: "the system is overwhelmed, overly bureaucratic, and hopelessly lost in lawyer land".  Readers might be surprised to hear that actually we lawyers don't live in a different land and, perhaps even more surprisingly, on the whole we tend to approve of increasing clarity in what we do (at least in the UK following the principles of the Woolf reforms).

This is one reason why, in contrast to many of the vocal individuals who opposed the old software patent directive that was shot down in 2005, we've tended not to dismiss outright the concept of reforming the principles around the grant of software patents (see various previous posts on Naked Law for examples).  As Lucy says: "Even though the European Parliament rejected the CII directive, that doesn't mean the status quo is OK. It just means that the CII directive was not the right update."

The concern from practitioners is that there is no uniform approach to patenting software across the EU (let alone in the rest of the world).  It is encouraging that Alison Brimelow, the new head of the EPO, recognises this: "What we need is not more patents, but more good patents".  IPKat has more information about Ms Brimelow here.

Comments

[Francis Irving]

Would you then support a proposed EU law to uniformly prevent EU member states from granting software patents, if there were one?

[Peter]

I'd probably describe myself as open to persuasion on economic and ethical grounds on the effect of software patenting generally - but find the arguments against quite persuasive (notably as expressed by Monsieur Lessig). But the tricky bit, as always, is in refining the detail. There will be some inventions that are clearly software and should not be patentable (but then there's already an exclusion from patentability for computer programs); there will be some that include software only incidentally and unless you're prepared to argue for the dismantling of the patent system generally - which I am not, and Gowers wasn't following his review - these should be patentable.

Which is where we get to the existing arguments about "technical contribution" and whether the invention lies in an excluded subject matter "as such". The status quo seems to me to be the regular granting of patents for software that should not be patentable, big discrepancies between EU nations over what is patentable, and a whole lot of uncertainty (see Richard's post today for an example).

I haven't worked out where the dividing line on the patentability of software should be drawn (I'll leave that one to the economists), but it strikes me that either codification of existing best practice or a fresh consideration of the issue would be a good thing.