AI and Europe’s medium-sized firms: How to overcome an Achilles heel

Artificial intelligence could become very real in the world of business. But funding needs to be corrected first for medium-sized firms, where millions of jobs are at stake.

Modern technologies collectively known as artificial intelligence (AI) have demonstrated their ability to help companies decide faster and operate better. With the promise of further and more impactful applications, business leaders in all sectors are starting to embrace AI technologies. This includes medium sized firms, which face particular challenges.

McKinsey estimates that the potential impact of AI in Europe could exceed 20% of gross value added in 2030, if widely diffused and properly used. Because of its enormous potential impact on the EU workforce and economy, AI has been embraced by public organisations and governments, generating a flurry of initiatives and large public budget allocations to support its spread.

Funding of AI adoption in Europe mainly comes from three sources. Public funding is the first source, either top-down or at the local level. The second is self-funding of large enterprises’ internal projects, which can in some cases be supported by public funding. And the third is venture capital funding of AI start-ups, totalling upwards of US$7.4 billion world wide in the second quarter of 2019 alone.

Funders, both public and private, operate under the assumption that investing at the top of value chains–such as Horizon 2020’s €1.5 billion investment in R&D–will ultimately benefit all of the economy–the famous “trickle-down” effect.

However, there are structural limits inherent to these funding mechanisms, creating a significant blind spot when it comes to medium-sized enterprises (MSEs). By MSEs, we mean companies employing between 250 and upwards of 5,000 staff and generating turnover between €50 and €6 billion. Current funding mechanisms exclude many MSEs, especially those whose primary activity is not centered on data, such as businesses that produce goods or services relying on the exploitation of a tangible asset. These MSEs are found in the manufacturing, transportation, and energy production sectors, among others. From our estimates, MSEs in these sectors (collectively referred to here as “industrial MSEs”) together represent about 20 million jobs in Europe, and 50 to 60% of gross value added.

MSEs do not benefit fully from these incentive mechanisms for three reasons. First, public support often comes in the form of financial incentives distributed as part of “top-down” programmes that operate via trickle-down mechanisms, all the way down to narrow segments of the industrial fabric. The groups targeted by these programmes often only feel their impact some time later. When this public support is delivered at the local level, its reach and scope are inherently limited by the amounts distributed, and by the distribution methods.

The second reason is that investments made by large corporations in AI technology only serve companies that are part of their value chain: suppliers, partners, and clients. However, not all MSEs are part of these value chains. A third reason is that venture capital investment is for the most part funnelled to products or platform start-ups in areas such as customer relationship management (CRM), or supply chain optimisation for instance. But medium-sized firms rarely adopt AI products that affect the core of their activity (for instance their manufacturing process), as doing so is often seen by their leadership as disruptive, and therefore both risky and costly.

Industrial MSEs are therefore overly exposed to the risk of a loss of competitiveness to foreign players, as industrial companies outside of Europe are moving fast in their AI transformation, acquiring flexibility, reducing costs, and expanding their global reach. For Europe’s industrial MSEs, job losses and value destruction for shareholders, on a large scale, is what is at stake.

To make matters worse, industrial MSEs in Europe encounter specific operational challenges when considering AI adoption. Due to legacy and the short-term constraints that come from running a mid-size business, not all MSEs have readily available data. They often struggle to understand what data to produce from their productive assets, and what to do with them.

There will be an acute shortage of AI skills on the market for the foreseeable future, especially the deep multi-disciplinary technological skills needed to tackle difficult industrial problems. It is an illusion to believe either that enough skilled AI experts will arrive on the market soon, or that most MSEs will be able to attract, let alone retain, the relevant talent.

Culturally, MSEs are often led by rather cautious leaders, who may hesitate to become early adopters in using new technologies, and may prefer to follow proven successes in their sector, especially when their core business or production assets are concerned.

To make progress, the current Europe-wide momentum around AI adoption, albeit admirable, must become more comprehensive. European governments should address this Achilles heel in the continent’s future economic leadership.

Solutions can be developed to fix this problem, and it is not necessary that these should all be “top-down”. At a minimum, Europe as a community of states needs to do three urgent things simultaneously.

First, it needs to increase AI awareness among the leaders of MSEs by engaging in aggressive AI education initiatives, and diffusing information on AI adoption success stories. Second, Europe needs to increase intra-EU market integration and co-operation around AI adoption by facilitating information exchange and access to AI solutions.

Last but not least, Europe should support the growth of pan-European “applied AI” labs, with a major role for private funding, that simplify and accelerate access to essential AI skills. To help industrial MSEs adopt AI solutions quickly, efficiently, and safely, such applied AI labs should combine deep, multi-disciplinary AI R&D skills with a pragmatic, proven culture of “getting things done”. These labs should also offer business transaction models that are compatible with the specific characteristics of MSEs and the constraints they face.

France is an interesting example in that respect. There are technology research institutes (IRT– Instituts de recherche en technologie) but only one of them, SystemX in Saclay, is in charge of helping companies with their digital transformation. And of course, AI is only a part of digital transformation. There is a scale issue that is linked to the granularity and diversity of SMEs. We need to enable the emergence of relays that are close to SMEs and ensure that their reach is wide enough.

The above actions are essential. They are required for the core fabric of the European economy-industrial MSEs-to be in a position to increase its performance and competitive standing in the global economy. Twenty million jobs are at stake.

References

AI4EU: https://www.ai4eu.eu/

Asgard and R. Berger (2018), Artificial Intelligence – A Strategy for European Start-ups”, https://asgard.vc/wp-content/uploads/2018/05/Artificial-Intelligence-Strategy-for-Europe-2018.pdf

CB Insights (2019), “AI in Numbers: Global Funding, Exits, And R&D Trends in Artificial Intelligence”, https://www.cbinsights.com/research/report/ai-in-numbers-q2-2019/

Council of the European Union (2019), “Draft Council Conclusions on the Coordinated Plan on the Development and Use of Artificial Intelligence Made in Europe”, https://data.consilium.europa.eu/doc/document/ST-6177-2019-INIT/en/pdf.

Deloitte (2019), “Future in the balance? How countries are pursuing an AI advantage”, https://www2.deloitte.com/content/dam/insights/us/articles/5189_Global-AI-survey/DI_AI-Global-Survey-Synopsis.pdf

European Council on Foreign Relations (2019), Machine politics: Europe and the AI revolution”, Policy brief, https://www.ecfr.eu/publications/summary/machine_politics_europe_and_the_ai_revolution.

McKinsey Global Institute (2019), ” Notes from the AI Frontier: Tackling Europe’s gap in digital and AI”, https://www.mckinsey.com/~/media/mckinsey/featured%20insights/artificial%20intelligence/tackling%20europes%20gap%20in%20digital%20and%20ai/mgi-tackling-europes-gap-in-digital-and-ai-feb-2019-vf.ashx

Région île de France (2019), « Intelligence artificielle : lancement du Pack IA et du Challenge ‘IA for Industry’ », https://www.iledefrance.fr/intelligence-artificielle-lancement-du-pack-ia-et-du-challenge-ia-industry

Dutton, T. (2018), “An Overview of National AI Strategies”, https://medium.com/politics-ai/an-overview-of-national-ai-strategies-2a70ec6edfd.

University of Cambridge (2019), “A survey of the European Union’s artificial intelligence ecosystem”, https://83d6fa69-2c07-4589-82a1-386547d3715c.filesusr.com/ugd/ff3afe_1513c6bf2d81400eac182642105d4d6f.pdf




Economic data

GDP growth: +0.5% Q2 2019 year-on-year
Consumer price inflation: 1.6% September 2019 annual
Trade: -1.9% exp, -0.9% imp, Q2 2019
Unemployment: 5.2% September 2019
Last update: 18 November 2019

OECD Observer Newsletter

Stay up-to-date with the latest news from the OECD by signing up for our e-newsletter :

Twitter feed

Subscribe now

<b>Subscribe now!</b>

Have the OECD Observer delivered
to your door



Edition Q2 2019

Previous editions

Don't miss

Most Popular Articles

NOTE: All signed articles in the OECD Observer express the opinions of the authors
and do not necessarily represent the official views of OECD member countries.

All rights reserved. OECD 2019