The state of AI defined by global adoption and regulation

Global AI adoption

A global survey of over 1,500 individuals across a variety of organizations, agencies and companies reported that by 2025, over 40% will implement AI in one or more of the seven patterns identified in the report. However, almost 90% respond that they will have some sort of in-progress AI implementation within the next two years. This would appear as a contradiction; however, if only 40% said they will implement AI by 2025 it would seem unlikely that nearly 90% would have some form of AI implementation in the next two years.

The most reasonable answer to this anomaly is that AI -- its uses and what technologies can be categorized as such -- is still misunderstood. This is an issue in long-term adoption of AI and work needs to be done to explain what AI is, how it differs from other approaches and how companies can go about implementing AI with the lowest overall risk.

The state of AI is still in the early majority phase, with many organizations waiting to implement the technology. The more organizations prove ROI and value from AI implementations, the greater the amount of companies that will follow, but until AI reaches broad adoption, we are in the educational phase of AI market growth.

When looking at global adoption patterns, the survey found little difference between the plans and implementation of adoption in regions across the world. North American and European firms are generally ahead in terms of pace of adoption, but Asian, Australian, African and South American companies are all showing vigorous interest in AI.

The report found that the biggest barrier to AI adoption is an insufficient quantity of quality data. This paired with the additional challenge of limited availability of AI talent and skills means that companies that are implementing AI are often slow to make progress.

For those firms that haven't started their projects at all, the primary challenges lie in justifying AI project ROI and the need to show how AI projects compare favorably to the traditional non-AI approaches to problems.

Country-wide AI strategies

Countries, not just companies, are coming up with plans and strategies to place their economies, militaries and governments in a competitive AI position. In Cognilytica's "Worldwide Country AI Strategies and Competitiveness" report, the research firm examines every country's AI strategy across the extent to which the country has a defined AI strategy, the level of government investments in AI, the extent of research and academic activity originating from the country's researchers and research institutions, the quantity of venture capital and entrepreneurial activity, and overall enterprise investment and implementation of AI systems.

France, Israel, the United Kingdom and the U.S. rate equally high at the top of the strategy list, according to the report, followed closely behind by China, Canada, Germany, Japan and South Korea. These latter countries are lagging a bit behind because they have relative weaknesses in individual areas.

For example, despite China's strong government financing, academic activity, venture capital and technology funding, and a well-established country-wide strategy, its largest enterprises are not investing in AI to the same extent as the leading countries.

Out of the top 15 ranked countries, 14 have very strong AI research activity. This shows there is a link between strong AI research and other quantifiable measures, such as venture funding, enterprise investment, government investment and country-wide strategy. However, the report finds that government funding, on its own, does not necessarily result in strength in these other areas.

Worldwide AI laws and regulations

Cognilytica also evaluated every country's overall legal posture, including laws and regulations, data privacy, lethal autonomous weapons systems, challenges relating to AI ethics and bias, among others.

Since, overall, AI adoption is in the early stages, it should be no surprise that most countries have not established any significant laws or regulations relevant to many of the areas above. However, many countries that have the strongest AI strategies are also starting to become more forward-thinking with regard to regulation.

The European Union has been the most active in proposing new rules and regulations, with existing or proposed rules in areas where regulation might be applicable to AI. The U.S., on both the federal and state levels, is also looking at many areas of AI laws and regulation, especially around the areas of autonomous vehicles, facial recognition and data privacy.

AI laws and regulations can be permissive or restrictive. In the case of autonomous vehicles, many countries and legislative regions are looking at loosening or making more permissive their vehicle laws to allow for autonomous vehicle operation on their roads. On the other hand, already 13 nations are looking at putting in restrictive laws to prevent the use or development of lethal autonomous weapons.

Many countries are seeing the challenges with widespread, unregulated storage and sharing of data and are implementing tighter data privacy rules and laws, which will have possibly unintended consequences with regard to AI. In general, it's hard for governments and legislative bodies to predict how such widespread technology like AI will impact society and their economies, and as such, it's quite likely that in the next 10 years we'll see laws emerge for situations we haven't even considered today.

These reports communicate the same message on the state of AI: that broad, widespread adoption of AI is underway across a wide range of organizations, use cases and countries. The use of AI has many countries not only building overall strategies to develop its growth, but also put in legislative and regulatory frameworks to deal with AI as it becomes an ingrained part of society.