Since the public release of ChatGPT in late 2022, artificial intelligence (AI) has been on everyone’s lips. With it, billions of people the world over suddenly woke up to the potential of generative AI.
The public debate around the opportunities and challenges of mass deployment of AI has snowballed. Meanwhile, businesses and organisations have been walking a fine line between excitement and concern. The excitement is in exploring strategic ways to incorporate AI in their dealings; however, the ultimate concern — for corporations and employees alike — is in staying relevant and preventing a full machine takeover.
Advertisement
The world of foreign direct investment (FDI) is no exception. While it’s still very early days, there are two main ways to think of AI in the context of FDI and investment promotion.
FDI into artificial intelligence
The opportunity
A first perspective relates to FDI into AI projects. Start-ups developing AI models have been scaling up very rapidly and it is only fair to assume they will continue to do so.
They raised $25bn in venture capital funding in the first six months of 2023, accounting for about one-fifth (18%) of global funding, Crunchbase figures show. “We are at the beginning of this,” says Gené Teare, the senior data editor of Crunchbase. Some of that capital is already being deployed to serve international expansion strategies.
AI has already become a prominent FDI force. Only in the research and development of AI-related applications, they have announced 778 projects worth $26.8bn since 2016, according to fDi Markets figures. India got the lion’s share, receiving 26.2% of that capital expenditure, followed by Canada (21.1%), Singapore (7%), the US (6.8%) and Israel (5.1%).
Advertisement
Tech behemoths such as Meta, Infosys, Amazon and IBM feature among the biggest foreign investors in AI-related R&D projects. German industrial group Robert Bosch is also among them, which is testament to AI being increasingly embedded in advanced manufacturing processes. It can also provide producers with services and servitisation of manufacturing.
Interestingly, AI is also proving a force in job creation as those projects created an estimated 160,172 jobs in high value-added functions, which maximises their knowledge transfer profile and their long-term development impact for host countries.
This is not a secondary detail: AI talent is as coveted as it is rare and concentrated in a few hotspots and companies. Overall, the percentage of AI practitioners does not exceed 1.1% of the total population of AI engineers in the US, 1.4% in Europe and 0.5% in China, according to data from Sequoia Capital. Anchoring the development of local AI ecosystems around big investors from abroad may well be a necessity, not just an opportunity for many countries.
The challenge
However, there is a flipside to the six-figure contract offered to high-end AI engineers. AI companies increasingly rely on closely monitored gig workers like data labellers, delivery drivers and content moderators. Perhaps millions of workers are recruited, particularly out of impoverished populations, to perform repetitive tasks often under precarious labour conditions and low wages.
AI may create jobs, but its jobs dividend has to be seen also through the lens of the quality those jobs it creates.
Besides, while there is no doubt it should be a priority sector for any geography, attracting FDI into AI is easier said than done.
If we look at it through the lens of contemporary geopolitics, not all AI is created equal. There is friendly AI and there is AI that is less so — and the roles swap across geographies.
Modern FDI screening frameworks have been designed to detect the difference and deter any AI-related FDI from companies and countries that raise national security concerns.
Beyond geopolitics, AI regulation (or deregulation) is already a major factor weighing on the minds of investors.
“It’s very important that we get the right balance between regulation versus innovation,” Vishal Marria, the CEO of British decision intelligence firm Quantexa, told fDi in July. On the one hand, “[too] much tight scrutiny ... stifles innovation”. On the other, “transparency in data lineage is critical to operationalising AI and machine learning when it comes down to a highly regulated market”.
While tightening regulatory restrictions to cross-border data flows limit the overseas expansion of AI companies, it can also indirectly encourage ‘barrier-hopping’ FDI to establish AI productive capacity and facilities in the individual foreign countries, so called market-seeking and data-seeking FDI.
Economic development organisations (EDOs) and investment promotion agencies (IPAs) have a key role to play in working at the intersection between the public and the private sectors, and advocating for rules that serve the internal security agenda while not deterring foreign AI companies altogether.
Artificial intelligence for investment promotion
The opportunity
EDOs and IPAs have a chance to raise their game through AI. AI-powered applications can enhance efforts across the whole spectrum of their functions — from promotion and facilitation to monitoring and aftercare.
As regards investment promotion, AI is a proven tool to boost marketing campaigns and digital outreach. AI can help with better investors targeting — for example through the screening and analysis of annual reports or quarterly earnings presentations. It can empower feasibility studies and support the development of bankable project pipelines.
When it comes to investment facilitation, AI can support the investor’s landing with things like online business registrations and opening digital bank accounts, as well as provide 24/7 support for any business-related inquiry. It can then feed the metadata from those inquiries back to the AI model to identify trends, common challenges and opportunities. As the domestic and international legal frameworks for cross-border investment become harder to read and keep up with, particularly in strategic industries, AI can also help investors with compliance.
Similarly, perhaps, to the early days of the internet, AI gives a chance for early adopters to gain competitive advantages over slower competitors. In particular, smaller organisations have a clear opportunity to punch above their weight with a wise, innovative incorporation of AI technologies.
The challenge
While the range of opportunities seems vast, AI is still a fledgling technology. As its use cases in and around FDI increase, playing it safe will become increasingly challenging.
First, AI models are not yet 100% reliable. “Though we believe that AI will be a game changer for investment promotion, we see that the technology has strong limitations,” Mr Grotto told fDi in August. He warns that IPAs “who go too fast in their adoption of AI might make some important mistakes”. In other words, going too fast may backfire.
Treatment of data also raises challenges. GAI applications such as ChatGPT have already stirred up much controversy around the way they crowd in data from all over the web. Considering the confidential nature of many conversations concerning future FDI projects, the treatment of any data captured by AI applications in this context is sensitive by nature.
Any IPA that is considering deploying GAI has to do so in a protected environment off the cloud, to make sure no confidential data is leaked.
Another broader challenge relates to the very raison d’etre of EDOs and IPAs, as the world embraces AI. These organisations are facing a Catch-22 situation. They have to test and deploy AI to stay relevant and ahead of competitors. But the more they do so, the more they rely on AI algorithms to find prospective investors, address their concerns and support them along the investment journey. And the less they may be needed in the form of fully fledged organisations.
At the same time, the same investors are incorporating AI in their decision-making processes. In a context where companies will have more access to data about specific locations, Pilar Madrigal, the director of investment advisory at Cinde, Costa Rica’s IPA argued in conversation with fDi earlier this year that the future trajectory of investment promotion can be boiled down to the following maxim: “The race gets faster.”
“It’s going to be harder for all of us to compete … All of us need to act quicker, more effectively and we’re not 100% prepared just yet,” she added.
While we can all agree that old fashioned in-person meetings and hand-shakes will continue to make a difference, the jury is out regarding the extent to which AI will make some jobs redundant — in investment promotion as well as any other sector.
Given the complexity of its system, regulatory environment and business landscape, AI is now suffering from a collective-action problem. As advanced technologies, including AI, can drive sustainable development, it represents a form of public good. This calls for mechanisms to address co-operation challenges.
At this stage, many of the questions that the mass deployment of AI raises are unanswered. What is clear is that this is only the dawn of the age of AI. Its uptake will soon accelerate, forcing everyone to adjust. By then, all organisations will have to be ready, lest they lose their one chance to make the most of the greatest disruption since the advent of the internet.
James Zhan is senior director for investment and enterprise at Unctad.
Jacopo Dettoni is the editor of fDi Intelligence.
This article first appeared in the October/November 2023 print edition of fDi Intelligence
More on artificial intelligence: