Software patents in Europe (presentation slides)

It’s all about the technical contribution, but what exactly is “technical” at the European Patent Office? Is it “technical” when engineers solve hard industrial problems with AI?

These and other questions were the topic of my presentation for the patent working group of Bayern Innovativ on 11 May 2023. It was a great honor to share my two cents after another great presentation by Martin Müller, Chairman of Technical Board of Appeal 3.5.06 at the European Patent Office.

These were the topics of my talk:

  1. Exponential times, anyone? Statistics on AI patent filings in Europe
  2. A walkthrough of the legal framework, EPO Guidelines and case law
  3. A hands-on patentability test of different hypothetical patent claims based on EP 4 000 000 A1.
  4. A checklist for structuring software patent claims
  5. A checklist for asking AI inventors the right questions

You can download my presentation slides (in German) on the “Downloads” page.

Hope it helps!
Bastian

2 comments

This site uses Akismet to reduce spam. Learn how your comment data is processed.

  • Hi Bastian,
    Signed up this morning ! With a sound biotech background rather a “newbee” in the digital world, but also in biotech AI etc. is making its way in… so high time to crank up our knowledge around it.
    Was looking with interest into your Bayern Innovativ presentation of 11 May 2023.
    And a little bit puzzled about slide 24, more specifically about the fact that non-trivial AI-features are not sufficient for a technical contribution ? “wobei das maschinelle Lernmodell (106) umfasst: “.
    In contrast, non-trivial processing of trivial features is indicated to be sufficient (slide 27) – makes sense.
    But would have thought that non-trivial AI features may be sufficient too as those may be important for the ML training ?
    Looking forward to your feedback.
    All the best,
    Franky

    • Hi Franky, thanks for your comment!

      I guess this goes back to the EPO’s notion that AI “as such” is only math, i.e., non-technical, and so it depends on what the AI is used for. For instance, if the AI is used to solve a business problem (think forecasting stock market developments), it makes no technical contribution even though the AI itself may be absolutely stunning. But if the AI is used to solve a technical problem (think controlling a manufacturing plant), it does make a technical contribution and therefore enters the inventive-step assessment.

      Makes sense?

      Best,
      Bastian