Why Half Of Companies Are Being Left Behind And How To Close The Gap

Thank you for reading my latest article The Two-Tier AI Economy: Why Half Of Companies Are Being Left Behind And How To Close The Gap. Here at LinkedIn and at Forbes I regularly write about management and technology trends.

To read my future articles simply join my network by clicking ‘Follow’. Feel free to connect with me via Twitter, Facebook, Instagram, Podcast or YouTube or subscribe to my other LinkedIn newsletters: AI & Future Tech Trends | The Intelligence Revolution | The Future of Work

Picture trying to run a marathon when you’re starting 46 kilometers behind your competitors. That’s the reality facing many companies worldwide right now as they attempt to scale AI while grappling with fragmented regulations, infrastructure gaps, and a critical shortage of digital skills. At the recent AWS GenAI & Data Day in London, I sat down with Sasha Rubel, GenAI Policy Lead for AWS in EMEA, and Alexandru Voica, Head of Public Affairs and Policy at Synthesia, to unpack this challenge and explore what it will take to close the gap between AI innovation and adoption.

The conversation revealed something unexpected. While everyone’s talking about the breakneck pace of AI development, the real story is the widening chasm between those transforming their businesses with AI and those stuck in pilot purgatory.

“AI adoption is actually outpacing the uptake of mobile phones in the early two thousands,” Rubel explained during our discussion. “That’s an extraordinary effervescence around the adoption that’s happening.”

Yet beneath this effervescence, a troubling pattern is emerging across three distinct dimensions. First, startups are steaming ahead while larger enterprises lag behind. Startups aren’t just using AI for productivity gains; they’re achieving what Rubel calls “strategic reinvention,” fundamentally reimagining how they serve customers and address their core purpose. Meanwhile, many large organizations remain trapped in the efficiency phase, unable to break through to operational transformation.

The second gap exists between the public and private sectors. This matters more than you might think. “When the public sector adopts AI, it actually inspires trust, which is one of the biggest blockers to AI adoption,” Rubel noted. More fundamentally, it allows governments to fulfill their core mission of serving citizens better.

The third divide separates tech-native companies from traditional industries like agriculture and manufacturing. This gap is particularly troubling because these non-tech sectors often have the most to gain from AI adoption. Rubel shared a compelling example of Italian farmer Mateo, who founded X Farm without any engineering background, using AI enabled by AWS to address sustainability challenges. His solution has now scaled across multiple countries, demonstrating what’s possible when we democratize access to these tools.

The most powerful AI applications aren’t necessarily the flashiest ones. During our conversation, both Rubel and Voica shared examples that illustrate the profound social impact of this technology when deployed thoughtfully.

Consider Hurone AI, a startup providing AI personalized treatment and advice for breast cancer patients in Rwanda. “Rwanda is a country where there’s one doctor per 3,200 women who are actively suffering from breast cancer,” Rubel explained. “There is no feasible way for that one doctor to see the 3,200 patients.” The solution isn’t replacing the doctor but augmenting their capacity to serve those most in need, with human oversight remaining central to the process.

Or take the work of HALO Trust in Ukraine, helping farmers like Mikael reclaim land contaminated by landmines. Within three months of the invasion, Mikael could no longer farm his land or feed his family. Using a mix of AWS AI and machine learning services, HALO Trust worked to demine the land, allowing Mikael to return to farming and support his community. “When we talk about ROI, we need to think not only in terms of solving the biggest challenges that face the world right now,” Rubel emphasized, “but also looking at the social transformation that this technology represents.”

Voica shared similar stories from Synthesia’s customer base. Synthesia is an AI video platform that allows organizations to create and distribute professional-quality videos at scale, including in multiple languages, without traditional video production. A small organization called Justdigit helps farmers in East Africa regreen land. They faced a challenge of training farmers across multiple languages on sustainable planting practices. “They used to record these training videos, but it took a lot of time and took a lot of effort,” Voica explained. By using Synthesia’s platform, they created training videos that could be distributed via mobile app, dramatically accelerating their impact.

These aren’t isolated examples. Genentech saved up to 43,000 hours of research time for drug discovery using generative AI on AWS. Using Synthesia, ServiceNow cut production costs by 50% for their academy training program, working toward an ambitious goal of training 3 million people.

If AI’s potential is so clear, why aren’t more organizations scaling it successfully? The answers emerged clearly during our discussion, and they go far beyond technology.

Infrastructure remains the foundation. Voica made a compelling comparison to railways and steam engines. “Right now we’re hyper-focused on models, benchmarks,” he noted. “But companies in general, from my experience, don’t really buy models. They want solutions and workflows.” Building bigger engines means nothing without the railway infrastructure to support them. This requires massive investment in data centers, cloud infrastructure, and computing capacity. Companies like Amazon are making significant commitments globally, including an £8 billion investment in UK AI and digital infrastructure, recognizing that without this foundation, innovation cannot scale.

Regulatory uncertainty creates paralysis. This is a global challenge manifesting differently across regions. Consider what’s happening in Europe as a cautionary tale: 68% of companies there currently don’t understand their obligations under the EU AI Act. The result? “Because of the lack of regulatory uncertainty, they actually intend to invest up to 30% less year on year in the technology,” Rubel revealed. These companies also spend up to 46% of their overall IT budget on compliance questions. The lesson applies everywhere: when companies can’t understand the rules or when regulations vary dramatically across markets, innovation stalls.

Rubel shared a historical parallel that brings this into sharp focus. In mid-19th-century London, when cars were first introduced, the Red Flags Act limited their speed to two kilometers per hour and required someone to walk in front waving a red flag. People walked instead because it was faster. “By the time the Red Flags Act was repealed, France and Germany had steamed ahead in their car industry,” Rubel explained. The lesson? Clear regulation based on shared definitions and real risk assessment accelerates innovation rather than hindering it. Overly cautious regulation does the opposite.

Funding gaps vary dramatically by region. While the conversation didn’t dive deep into specific numbers, both Rubel and Voica emphasized the disparities in venture capital investment across different regions. In some markets, startups struggle to find the capital they need to scale AI solutions, while in others, like the U.S. and parts of Asia, funding flows more freely. Startups need clear pathways to capital, and many identify access to funding as critical for faster growth. The companies that can secure investment move faster, widening the gap further.

The skills shortage remains the most fundamental barrier. “One of the things that is the biggest blocker to AI adoption right now is a lack of digital skills,” Rubel stated plainly. This is a universal challenge. In some markets, it can take up to six months to find someone with the necessary digital skills. Six months in AI years, as Rubel put it, is far longer than dog years given the pace of technological development.

What’s fascinating is who wants to learn. Research shows the demographic most interested in learning digital skills is people over 60, those closest to retirement. “People over 60 understand that their capacity to participate in the world of tomorrow will actually depend on their capacity to understand and proactively use these tools,” Rubel explained. Yet when researchers asked employees across different markets what was holding them back, the answer was consistent: they don’t have time, and classes are expensive.

One of the most actionable insights from our conversation came from Voica’s discussion of technical standards. Synthesia became the first generative AI company certified with the ISO 42001 standard, designed to help organizations manage AI risks. The impact was immediate and practical. Previously, conversations with large organizations involved lengthy back-and-forth between legal teams about compliance, governance, and risk, sometimes lasting months. “Now that we are ISO 42001 certified, that means that those conversations have been rapidly accelerated,” Voica said. “We can skip through to the commercial conversation.”

This illustrates Rubel’s broader point about how responsibility and innovation must go hand in hand. “There’s a big debate right now in the policy space that’s very counterproductive, which posits responsibility and innovation as opposites,” she noted. “Actually, responsibility drives the trust, which is a huge blocker to AI adoption, which will drive the adoption, which will fundamentally drive the innovation.”

Looking ahead, the conversation naturally turned to agentic AI, the next wave of automation that’s already showing concrete results. Rubel defined it as “an AI system that, with increased autonomy, deploys and executes tasks, whether that’s through API calls or else, and that also actually increases its accuracy through feedback that allows it to improve its results.”

Voica added: “… agentic AI is like having a country of subject matter experts in a data center… It’s not just about getting answers from them, but getting them to do work as well,” he explained.

The applications are already impressive. Mindflow, a French startup, is deploying agentic AI run on AWS inside large enterprises to address cybersecurity threats, automating responses to attacks that happen every second. Synthesia is building video agents that combine AI avatars with stacked models, creating interactive, personalized learning experiences. Another startup, FlowX.AI, has built agents in financial services that process mortgage applications and assess fraud risk while maintaining the governance and transparency banks require.

What excites Rubel most is how this technology augments human creativity. Agentic AI frees up time for us to focus on what she calls the three C’s: creativity, curiosity, and critical thinking.

Perhaps the most important question is how this technology will reshape work itself. There’s understandable anxiety about job displacement, but history offers perspective. When the printing press was invented, monks who had spent their lives painstakingly copying texts by hand broke the early machines because they feared losing their purpose. Yet the printing press created entirely new industries: publishing, editing, independent authorship, and the vast field of literature we enjoy today.

“Jobs disappear, but in the long run, more jobs are created,” Rubel noted. “The fundamental question is, how do we prepare individuals today for the world of work of tomorrow? And that’s through digital skills.”

Here’s what makes this moment different from previous technological shifts: the tools are becoming radically more accessible. “The beauty is that nowadays it’s no longer technical,” as I pointed out in our conversation. You can speak to AI to create code, develop solutions, and handle tasks that previously required specialized training. The barrier to entry is collapsing.

Voica shared that when Synthesia launched eight years ago, they expected video editing professionals to be their primary users. Instead, they discovered people who had never made a video in their lives were adopting the platform. “That for us was the key to success, to build a platform that anyone could use,” Voica explained. Today, 90% of the Fortune 100 use Synthesia, up from 47% just two years ago.

The future of work, as both Rubel and Voica envision it, is one where AI handles the tasks nobody wants to do anyway. “Nobody wants to fill in paperwork,” Rubel said, “and that actually returns us and gives us back time to do what makes us fundamentally human, which is the creative, constructive, collaborative possibility of us coming together.”

Voica added an important dimension often overlooked in technology discussions: “What I enjoy about work is not just acquiring new skills and applying them, but forming connections with the people that I work with.” If AI can help us learn faster and be more productive, we’ll have more time for the fundamentally human aspects of work, the collaboration and connection that make it meaningful.

The path from AI innovation to adoption isn’t a technology problem; it’s a systems problem requiring coordinated action across multiple fronts. We need infrastructure investment, regulatory clarity with international interoperability, clear pathways to funding, and accessible digital skills training. Most fundamentally, we need to foster what Rubel calls “a culture of why,” starting from business challenges and working backward to solutions.

The encouraging news is that the tools are here, the use cases are proven, and the ROI is clear. Organizations like Formula One are using AI run on AWS for real-time analysis during live races. The BBC is unlocking 25 petabytes of archived content. WPP is building 60,000 personal agents across its organization. These aren’t hypothetical futures; they’re happening right now.

The question isn’t whether AI will transform how we work and live. It’s whether we’ll create the conditions for everyone to participate in that transformation. As Rubel emphasized, “We have a tech sector as diverse as the societies that we aspire to create.” That means making digital skills freely available, building trust through responsible deployment, and ensuring the benefits flow broadly rather than concentrating in a narrow tier of early adopters.

The gap between innovation and adoption isn’t inevitable. It’s a choice. And the organizations, governments, and individuals who close that gap will be the ones shaping the future rather than scrambling to catch up to it.

To learn more about Generative AI at AWS, click here: https://aws.amazon.com/ai/generative-ai/

#AWSPartnership #Sponsored

Bernard Marr is a world-renowned futurist, influencer and thought leader in the fields of business and technology, with a passion for using technology for the good of humanity. He is a best-selling author of over 20 books, writes a regular column for Forbes and advises and coaches many of the world’s best-known organisations.

He has a combined following of over 5 million people across his social media channels and newsletters and was ranked by LinkedIn as one of the top 5 business influencers in the world. Bernard’s latest book is ‘Generative AI in Practice’.