IBM dominates AI patents, startup CognitiveScale penetrates the top 10

]Artificial Intelligence is an emerging transformative technology that is poised to revolutionize the way we live. The race is on to build AI systems that successfully simulate human intelligence. Companies have begun to recognize the advantages AI can bring into their products and have started pouring investments into research & development (R&D) efforts in the field, investing in larger compute power, and acquiring talent from universities and start-up companies. Notably, the increasing adoption of AI has resulted into a predicted significant growth increase in the AI market, with market growth rate estimates reaching as high as 60% for 2017-2022 [1].

Based on our analysis, IBM has had more AI-related utility patents and patent applications (P/PA) with filing dates over 2004-2018 than any other company. Fig. 1 & 2 highlights how dominant IBM is in AI. Microsoft, which follows IBM, only reached 26% of IBM’s total P/PA count.

IBM has shown commitment in developing their research and development arm, spending about 7-8% of their annual revenue in R&D in 2017 [2]. IBM also tied-up with MIT to focus on fundamental AI research in AI Algorithms, the Physics of AI, the Application of AI to industries, and Advancing shared prosperity through AI [3]. One of IBM’s most notable AI products is Watson, a supercomputer configured to perform as a question-answering machine. IBM Watson is powered by machine learning which integrates AI to different applications [4].

Considering only P/PA with the filing date 2013-2018 (Fig. 2), the same companies are on the top 3, with IBM again leading by a large margin. IBM has held the number one position in terms of most U.S. patents granted for the past 25 years [5]. IBM recently highlighted all its efforts in AI in Think 2018, IBM’s combined partner, customer, and developer conference [6].

One of the surprising companies to penetrate the list of top assignees is CognitiveScale at 7^th place for P/PA with filing date 2004-2018 and 4^th place for 2013-2018. CognitiveScale, a startup founded only in 2013, is offering cognitive cloud platform which mines data from very large databases. One of its products is “Augmented Intelligence” software which emulates and extends human cognitive function through the pairing of people and machines, which CognitiveScale claims as the most effective way to maximize the value of AI [7]. The concept of augmented intelligence is also becoming attractive in industry-specific applications. In the banking industry, for example, augmented intelligence enables banks to apply AI to gathering and analyzing massive amounts of data, which helps ensure bank employees have the right answers and advice for their clients [8].

Apart from Amazon, which is mainly an e-commerce company, all the top assignees are mostly tech companies which provide software/IT services and/or manufacture electronic devices. Tech giants Microsoft, Google, Facebook, and Intel are known to actively pursue research in AI and infuse said technology into their respective product experiences. NEC has developed NEC the WISE, a portfolio of AI technologies for enriching human intellect and creativity. Oracle and SAP are also enhancing their SaaS applications by incorporating AI and machine learning. Samsung envisions all its devices will have baked-in AI by 2020 [9]. Fujitsu developed Deep Tensor that can directly analyze the relations among numerous pieces of real-world data, as well as technology that enables AI to explain basis for its judgment from input to output [10]. Amazon uses AI in product recommendation, shipping routes, Amazon Web Services, and its virtual assistant, Alexa.

Fig. 3 below shows the P/PA distribution per year of the top assignees for the filing date 2004-2018. Based on Table 1, CognitiveScale and IBM showed the most significant increase in P/PA filings during the last 5 years or so (2013-2018) compared to the entire 2004-2018 period.

 Table 1. Percentage of P/PA from 2013-2018 relative to total P/PA over entire 2004-2018.