AI predictions for 2026: keynote from Benoit Reillier

The past year in AI is a tale of technological and social change at pace – perhaps of a speed we’ve not seen before. This is only poised to continue next year and beyond. “If you feel like things are accelerating around you and innovation is getting quicker, that’s because it is,” affirms Benoit Reillier, co-founder and CEO of Launchworks & Co.

This leaves regulators and practitioners with a shared question: How do we regulate and build products in a world we cannot easily predict?

The strategic importance of ‘surfing the waves’

AI is the latest wave in a much longer cycle of innovation that started with the industrial revolution. From steam power and electricity to electronics and digital networks, each wave has reshaped economies and redistributed competitive advantages. What makes the current moment distinct is, again, its speed.

“If you feel like things are accelerating around you and innovation is getting quicker, that’s because it is.”

“These cycles are getting shorter and shorter,” says Benoit. “It’s absolutely critical to seize these waves of innovations.” The digital wave of the past 30 years has widened economic gaps between countries. In 1990, GDP per capita in the US and Europe was broadly comparable. Today, the US has pulled decisively ahead, with numbers nearly double those of the UK. A key driver has been the growth of digital platforms able to orchestrate global ecosystems rather than simply sell products. “Today, the largest companies in the world are these digital platforms: very successful businesses, attracting participants, connecting them, matching them and enabling them to transact globally,” Benoit explains.

That concentration of power is now shifting again. The world’s most valuable companies increasingly include AI infrastructure and semiconductor firms, signalling that strategic advantage is moving deeper into the technological stack.

Looking at AI within a historical context, Benoit identifies three macro trends that will shape the next decade and how each will start to play out in 2026. As the next wave gathers strength, firms, platforms and institutions need to ready themselves to adapt in order to succeed.

1. Scaled cognition

Once a finite, human resource, intelligence is becoming abundantly accessible for the first time. Advanced AI systems support knowledge work, but they also replicate and scale it.

Agentic AI

As we look ahead, agentic AI will continue to be a cornerstone of scaled cognition. “This process has already started, but we think it’ll accelerate in 2026,” says Benoit. “Increasingly, you’re going to be able to delegate things to agents.” AI agents will work in coordinated systems, orchestrating workflows across functions and platforms.

Many organisations are already preparing for this shift. “Just under 50% of Fortune 500 companies said that they would have some of their core processes running with agentic AI next year,” Benoit notes. While this promises efficiency gains, he cautions that business strategy must look beyond optimising productivity in order to stay competitive.

“Thinking strategically about what is going to change in your ecosystem as a result of AI, and how that impacts the nature of what your organisation is going to have to do to add value, is one of the most interesting discussion we’re having with our clients”

AI-mediated discovery and commerce

AI is also transforming how users discover and access information. Search is becoming conversational, contextual and increasingly delegated to AI agents. Users no longer rely on keywords but describe goals, preferences and constraints in dialogue with their AI tool of choice.

“Organisations really need to develop new skills and capabilities to i) find what they’re looking for and ii) be discoverable online.”

At the same time, the supply of information is changing just as dramatically. A rapidly growing share of online content is now also generated by AI, with 90% of web content expected to be AI-generated by 2026. In such an environment, Benoit thinks, visibility and trust become harder to secure: “Organisations really need to develop new skills and capabilities to i) find what they’re looking for and ii) be discoverable online.” This important topic was also discussed at the event during a dedicated panel.

2. Physical & digital convergence

If scaled cognition changes how we think, physical-digital convergence changes how we live. From home devices to self-driving vehicles and humanoid robots, AI is beginning to move into our everyday environments.

Incarnated AI

“AI is going to be embedded in lots of physical things around us: glasses, mobile phones, your hoover or your [virtual] pets,” says Benoit. While early-generation robots will often have limited practical value, their symbolic impact will be significant. By the end of the decade, second- and third-generation incarnated AI could become genuinely useful in many households. And almost a fourth of event participants on the day said they thought they’d have a humanoid robot at home by 2030.

Phygital fusion

Incarnated AI is further enabled by advances in digital twins and simulated environments. Thanks to advancements in spatial computing, these “mirror worlds” are more and more sophisticated. Robots can now be virtually trained millions of times faster than through physical trial and error, which dramatically accelerates development. 

Meanwhile, digital systems are already shaping real-world behaviour. Navigation platforms, for example, no longer merely analyse traffic patterns, as Benoit illustrates: “If Google Maps decides to make you believe that the A4 is closed, the entire traffic in London will change immediately. Whether or not the A4 is closed is almost irrelevant – the AI will reroute everybody as if it were. It’s having an impact on the [physical] world.”

3. Governance & alignment

As AI systems scale and ecosystems grow more complex, aligning innovation with human needs – from safety and fairness to sustainability and trust – will demand more nimble and adaptive institutions. Strong AI alignment is becoming a practical necessity.

Open ecosystems

Regulators are pushing for greater openness, interoperability and accountability across digital ecosystems. “Apple, which notoriously likes to keep its ecosystem quite closed, [is] being forced to open up its app store, its entire ecosystem, its hardware, et cetera, to third parties,” observes Benoit, noting that this change is in compliance with the EU’s Digital Markets Act. “This is going to play out in 2026,” he adds. This poses major challenges for some firms, but also presents potential opportunities since well designed open ecosystems can foster growth while being surprisingly resilient.

Responsible tech

Some organisations are beginning to explore regenerative and net-positive models, recognising that, as Benoit puts it: “It’s not enough to do net zero.”

“Tech alignment is really the call for human centricity to be embedded in the way we design and use technology.”

Beyond sustainability, Benoit reemphasises the theme of alignment. “Tech alignment is really the call for human centricity to be embedded in the way we design and use technology,” he says. This is particularly crucial given the anxiety and uncertainty that many feel about AI and how it might disrupt the labour market and displace jobs: “Governments are starting to see that and trying to see how they can create an environment that has positive outcomes for everybody.”

Join the conversation

This keynote from Benoit Reillier took place at Platform Leaders, hosted by Launchworks & Co on the 18th of November 2025 in London. Visit the Platform Leaders website to watch session recordings, read more articles, join the Platform Leaders community and continue the conversation. Explore the agenda, watch session recordings and join the community to stay up to date with future events and insights. To watch the full event, you play the video below.