Expert Predictions For DeepTech In 2026, Part 1

Deeptech is entering a decisive new phase where breakthroughs are no longer confined to research labs and pilot programmes. As 2026 approaches, industry leaders are expecting the convergence of AI, advanced computing, decentralised systems and novel engineering to reshape entire sectors at once – and rapidly.

Although it’s impossible to know for sure, it seems as though the coming year is set to challenge long-held assumptions about what technology can achieve, who can build it and how quickly it can scale. These predictions from experts across the field outline the themes that will define deeptech’s next wave and the opportunities founders should prepare for now.

 

Intelligent Systems Move from Assistance to Autonomy

 

Indeed, it seems as though 2026 may be the year autonomous systems truly become independent operators rather than sophisticated assistants. AI agents will increasingly manage complex workflows, negotiate on behalf of users and make context-aware decisions without human prompting.

This shift, many believe, will be enabled by rapid advances in verifiable data, new trust frameworks and hardware designed specifically for on-device intelligence. The result should be safer, more reliable autonomous systems that businesses can trust in high-stakes environments. From manufacturing floors to financial modelling, autonomy will stop being experimental and start becoming routine – it’ll be part of our everyday. Autonomy will be normal.

 

The Infrastructure Layer Gets Rebuilt for Trust and Verification

 

Deeptech leaders agree that the next frontier is not more models or more data – rather, it’s confidence in the provenance, accuracy and security of what systems create.

Next, year, trust infrastructure will become the new backbone of innovation. Companies will invest heavily in technologies that verify AI-generated content, authenticate device identities and prove the integrity of data flows.

This is expected to create new markets for verifiable digital credentials, cryptographic tooling and decentralised infrastructure. And, as businesses begin to rely more heavily on autonomous tools, the demand for certainty will almost certainly reshape product design, regulation and the competitive landscape.

 

 

Our Experts

 

• Vera Kunz: Investment Associate at APEX Ventures
• Rajiv Ramaswami: CEO of Nutanix
• Jarrod Vawdrey: Field Chief Data Scientist at Domino Data Lab
• Gareth Cummings: CEO at eDesk
• Tony Gentilcore: Co-Founder, Engineering at Glean
• Jimmy Tam: CEO of Peer Software
• Jan Ursi: VP Global Channels at Keepit
• Dr Irina Babina: CEO of Concr
• Xiangpeng Wan: Product Lead and Strategic Research Lead at NetMind.AI
• Sabrina Maniscalco: CEO and Founder of Algorithmiq
• William Stevens: Co-Founder of TechTour and Lead Coordinator at EIC Scaling Club’s
• Larissa Schneider: COO and Cofounder of Unframe AI
• Brady Lewis: Senior Director of AI Innovation at Marketri
• Dr. Tatyana Mamut: Co-Founder and CEO of Wayfound

 

Vera Kunz, Investment Associate at APEX Ventures

 

 

“The deep tech investment landscape is shifting fundamentally in 2026. We’re moving away from the ‘bigger is better’ obsession with massive AI models toward what we call ‘intelligence density’—deploying expert-level reasoning on smaller footprints with zero latency. Edge AI and specialized models will dominate, driven by unsustainable cloud inference costs and enterprise demands for data sovereignty.

“This efficiency trend connects directly to sustainability, becoming a performance driver rather than a standalone impact category. Deep tech areas like quantum, advanced compute, and materials discovery are increasingly recognized for delivering superior unit economics that happen to reduce energy consumption and emissions. Capital has become more disciplined—we’re prioritizing technologies with credible scale-up paths.

“The most compelling opportunities in 2026 lie at the intersection of efficiency, security, and sustainability. Startups delivering private-by-design architectures with minimal energy footprints will capture significant attention from both investors and enterprises seeking resilient, resource-efficient solutions.”

 

Rajiv Ramaswami, CEO of Nutanix 

 

 

“The sovereign edge will continue to evolve.

“AI is a force for more distributed infrastructure as AI moves out to process data generated at the edge. Enterprises will need to consider the global management, distributed security, and remote recovery/destruction policies available for the sovereign edge and rely more on platform engineering to successfully achieve this.

“As AI continues to skyrocket in adoption, businesses will look to find ways to process AI-related data locally. As a result, organisations will look to global management solutions with integrated security and edge resiliency to help keep this in check.”

 

For any questions, comments or features, please contact us directly.

 

Jarrod Vawdrey, Field Chief Data Scientist at Domino Data Lab

 

 

“2026 is when the music stops. CFOs are done writing blank checks for “AI innovation” that can’t be tied to actual business results. We’re already seeing enterprises start to pump the brakes on a significant percentage of their planned AI spending because leadership finally asked the obvious question: “What are we actually getting for this?” And most teams have no good answer from a year of PoCs that never made it into production.

“The handful of use cases that actually move numbers will survive. Revenue up, costs down, cycle time reduced… real KPIs that matter. Everything else gets killed. No more pilot purgatory, no more “let’s experiment and see,” no more demos that wow executives but never ship. If you can’t show business impact in three to six months, you’re done. The companies winning in late 2026 are the ones who got religious about measurement early and weren’t afraid to kill their darlings.”

 

Gareth Cummings, CEO at eDesk

 

 

“In 2026, a meaningful share of customer interactions will happen agent-to-agent. Shoppers will use AI assistants to check stock, confirm delivery times or verify returns, and brands will respond with their own AI agents that can read order data and act instantly.

“The real shift is speed. Conversations that used to take minutes will collapse into a single automated exchange. For e-commerce, this will separate retailers with unified data from those still stitched together with fragmented systems. The first group will meet machine customers effortlessly. The second won’t be able to participate.”

 

Tony Gentilcore, Co-Founder and Engineering at Glean

 

 

“2026 marks the start of natural selection for automation. Tools that create noise instead of value will go extinct and the AI bubble will burst. The ones that thrive will be grounded in business context and real outcomes. The age of novelty is ending as evolution now favours ROI.”

 

Jimmy Tam, CEO of Peer Software

 

 

“Agentic AI will converge with distributed file services to enable a new class of distributed digital teams.

“2026 will mark the beginning of a major architectural shift: agentic AI systems will merge with distributed file services to create AI digital teams that can autonomously capture data, act on it, and push results across multiple locations and platforms. As organisations deploy distributed AI agents at the edge, in the cloud, and across data centres, they will realise the missing piece is the ability to move information seamlessly and intelligently between those agents.

“The convergence of agentic AI and distributed file services will become essential for orchestrating workflows, sharing context, and ensuring AI agents can collaborate in real time across heterogeneous environments.”

 

For any questions, comments or features, please contact us directly.

 

Jan Ursi, VP Global Channels at Keepit

 

 

“By 2026, compliance expectations will become embedded in nearly every SaaS data protection RFP. Requirements tied to NIS2 and DORA will shift from “requested” to “assumed,” especially in finance, energy, healthcare, and the public sector. Organisations will insist on local digital sovereignty: data stored in-region, zero sub-processors, and guaranteed access even if the original SaaS platform is unavailable.

“Because many companies are still in the early stages of meeting these regulations, demand will rise sharply as deadlines tighten. Local partners will play an essential role. They understand national sovereignty rules, infrastructure constraints, and the operational realities of regulated industries. As a result, the channel will become a core enabler of compliant SaaS adoption, not an afterthought.”

 

Dr Irina Babina, CEO of Concr

 

 

“In 2026, we’ll finally move from talking about the potential of AI in healthcare to seeing real‑world impact. Right now, most approved tools are still relatively narrow in scope (think imaging algorithms for radiology or digital pathology), and they’re used more to analyse data than to generate real clinical insight. That will change as AI systems start to inform decisions across the full drug development and treatment lifecycle, especially in research settings where regulation is more flexible.

“The most exciting developments will sit at the intersection of biology and technology, where the biggest gains for humanity lie. Digital twins are a good example: they model how individual patients might respond to treatments using vast multimodal datasets, which cuts risk and speeds up development timelines in ways that weren’t possible before.

“At the same time, healthcare tech is bypassing the system entirely and going directly to patients. The NHS working with Neko Health shows where this is heading. But direct-to-consumer models could exacerbate issues around trust, affordability and accessibility.

“This is also the most promising time for biotech startups in years. Big pharma is hitting a patent cliff in 2027 as its blockbuster drugs lose exclusivity, opening the door for competition. That will drive them to hunt for smaller groups with promising treatments, often quite early in development. That’s where deeptech tools in particular could come in handy, modelling drug potential in the market, projecting population-scale effects, and designing model-informed strategies, which regulators like FDA are encouraging.”

 

Xiangpeng Wan, Product Lead and Strategic Research Lead at NetMind.AI

 

 

“The premium on closed AI models will collapse in 2026. Airbnb’s CEO already proved this when they shifted massive workloads to open-source models like Qwen to cut costs. The gap between expensive closed models and free open ones has essentially disappeared. We’ve hit a physical wall in reasoning capabilities.

“Companies will also no longer be willing to let their core data sit in black-box clouds anymore. Data sovereignty means businesses are moving AI back to their own infrastructure.

“Meanwhile, the bottleneck in the AI arms race is shifting from silicon to electricity. Data centres have the GPUs but can’t plug them in. Power grids can’t keep pace with training, reasoning and inference demands. Microsoft’s future isn’t constrained by NVIDIA anymore; it’s constrained by nuclear plants and substations.

“In the workplace, job adverts for junior roles will diminish. Companies won’t hire and train expensive newcomers when one architect can manage AI agents. Any job without high-level judgment or genuine human connection is getting automated to save money.”

 

For any questions, comments or features, please contact us directly.

 

Sabrina Maniscalco, CEO and Founder of Algorithmiq

 

 

“For years, the quantum industry has been marked by isolated claims of ‘supremacy’ and ‘advantage,’ often driven by headlines rather than verifiable science. This year, that changed. With IBM, our team achieved and independently verified a real quantum advantage, setting a new benchmark for transparency in the field.”

“Quantum advantage is the point where a quantum computer outperforms any classical system on a meaningful task. Now that this threshold has been crossed, the question shifts from whether quantum can deliver to where it will have the biggest impact first. Drug discovery and life sciences are leading contenders, as classical supercomputers face fundamental limits in simulating nature at high accuracy.”

“Just as important is how we reached this milestone. Our work with IBM introduced the first quantum advantage tracker, helping unify the community around open, rigorous, and repeatable verification. It marks the beginning of a new era where progress is defined cooperatively, through shared, scientifically validated results.”

 

William Stevens, Co-Founder of TechTour and Lead Coordinator at EIC Scaling Club

 

 

“Next year will be about capacity building in defence. There is a big gap of applied software and AI talent for dual use and defence applications as European defence forces are building their resources, innovation base, and sovereignty. It will generally be a game changer for deep tech with space, photonics, AI for the edge and autonomous systems, quantum and cybersecurity growing significantly – cross-national cooperation and procurement is, however, a big pain point.”

 

Larissa Schneider, COO and Co-Founder of Unframe AI 

 

 

“Full-stack AI companies will start popping up

– A full-stack AI company doesn’t just sell tools. It builds an entire business powered by AI and competes directly with incumbents.
– Owning the full workflow becomes the moat.
-Legacy players will need to speed up their AI adoption or risk getting overtaken by AI-native challengers that move faster and operate leaner.

The 10-year AI strategy era is over

– Enterprises that spent years building internal AI platforms will realize those efforts didn’t produce meaningful outcomes.
– They’ll turn to external partners who can deliver business outcomes fast and at scale, while keeping only their real core IP in-house. Building the entire AI stack themselves doesn’t create value and pulls focus away from what actually moves the business forward.

The ‘quick DIY’ wave will fade out

– Fast, no-code app generators look attractive and generated a lot of buzz, but most outputs aren’t production-ready, let alone enterprise-grade. Teams try them, quickly reach their limits, and move on.
– Loveable is a good example—great for prototypes, not for production.
– They’re even hiring FDEs now because the pure DIY motion isn’t working. Enterprises need reliability, custom integration, accuracy, and security, which simple DIY tools can’t deliver at scale.

LLM progress will slow down

– Models will keep improving, but only incrementally. There’s no big step change on the horizon.
– That shifts the real advantage to the application layer and using what we already have to solve real-world problems, not waiting for the next breakthrough.

Outcome-based pricing becomes a priority

– Enterprises are tired of AI projects failing at a 95% rate, so they’re pushing for pricing tied to real business outcomes instead of raw consumption.”

 

For any questions, comments or features, please contact us directly.

 

Brady Lewis, Senior Director of AI Innovation at Marketri

 

 

“In 2026, deep tech will no longer just be about the discoveries themselves (as it was historically). Instead, it will become part of a company’s day-to-day operations. Mid-sized businesses will be able to access the established foundational AI Models.

“These businesses will create the environment and opportunity for the use of deep tech through all areas of their business, and mid-sized businesses will also have access to teams of applied intelligence that will assist them in integrating AI, automating processes, and using data-driven reasoning for both their internal and external business practices.

“Companies will move from merely testing AI-based solutions to completely redesigning their existing processes to support and accommodate the integration of AI-based solutions into their businesses.”

 

Dr. Tatyana Mamut, Co-Founder and CEO of Wayfound

 

 

“If companies want to deploy AI agents in 2026 without being left behind, they need to stop treating agents like traditional software. We’ve moved past experimentation into real deployment, and agents behave more like humans than code; they learn, they adapt, they make autonomous decisions, all of which require constant supervision.

“Right now, different departments are building their own agents with zero central oversight. Companies finalizing their 2026 budgets need to understand that if they don’t budget for AI agent supervision, they’re walking into sprawl, compliance nightmares, and zero visibility into whether their agents are even working correctly.

“The companies that will win are the ones treating AI agents like team members who need performance reviews, training, and active management. Company messaging changes all the time, and someone has to communicate that to your agents. Agents drift from goals just like employees do, which means you need systems to catch it as it happens. This is why we built Wayfound: to give leaders centralized oversight and supervision across all their agents, no matter what platform they’re on. Competitive advantage in 2026 comes down to having AI agents you can actually manage and trust to work alongside your people.”

 

For any questions, comments or features, please contact us directly.