What’s next: IRIS outlook for 2026
What’s next: IRIS outlook for 2026
“If we want things to stay as they are, things will have to change.” Archibald Colquhoun
A few weeks into 2026, clearer patterns are beginning to emerge across the tech landscape. Not because uncertainty has faded, but because the system’s pressure points are now coming into view.
Get a glimpse into how Julien-David Nitlech, Managing Partner at IRIS, sees the year likely to unfold. His views also feature alongside peers in The French Tech Journal - French VC Outlook 2026: More concentration, fewer bets, bigger stakes with analysis by Helen O’Reilly-Durand.
Weirdest year ever... until 2027?
2026 opens with a strange sense of déjà vu. The cadence looks familiar: AI mega-fundraises, a race for exposure to AI or defense, capital clustering around a narrow set of narratives. But beneath the surface, the ground is shifting.
There's two structural AI forces are at work.
- First, AI leadership is being internalized. Google, via Gemini, has absorbed the full stack. Others remain structurally dependent on external capital and partnerships that once felt permanent, until they weren’t.
- Second, Chinese AI is driving the marginal cost of “AI as a feature” toward zero, much like cloud did between 2008 and 2010.
The implication is uncomfortable. A large share of current AI valuations assumes linear progress and durable pricing power. A collective realization that “it is complicated” would not be marginal. It would ripple through ongoing and future rounds, with potentially violent effects.
Industrial AI and the infrastructure moat
The AI bull run continues, but its nature is changing. Multimodal and agentic systems are expanding their operational footprint across enterprises. Upstream, a clear premium is emerging around data readiness. Tools that flatten, orchestrate, and industrialize data flows are becoming mission-critical. This is driving valuation tailwinds for data centers, data lakes, and the broader data infrastructure stack.
Hyperscalers continue to consolidate aggressively, using M&A to assemble end-to-end, AI-driven customer experiences rather than isolated features. At the growth stage, companies with deep product penetration are widening their moats by embedding AI directly into existing client bases, particularly in health, manufacturing, and industrial technology.
When AI cheapens, value moves elsewhere
As AI becomes increasingly abundant, its value will relocate.It will move toward execution and toward companies that get things done. Robotics players and industrial production companies operating under intense margin pressure are likely to drive the next wave of adoption.
Software, robotics, and applied AI embedded directly into real operations will matter far more than standalone intelligence. Amazon has already shown us the blueprint. 2026 will surface early signals with real volume likely to be seen in the following years.
After defense, space becomes strategic for Europe
Defense has repriced sovereignty and security, but differentiation and exits are increasingly hard to read. Drone systems and AI-native defense platforms mature quickly, often faster than capital markets can price them. The next strategic layer is space.
Owning payloads, private access to data, connectivity, and imagery will matter in the same way owning cloud or proprietary data once did. Space is shifting from state-led infrastructure to early commercial trade. The first true platforms are now emerging.
Social systems will not reform. They will be bypassed.
Healthcare, education, and retirement systems are structurally unaffordable, and governments are unable to move fast enough to adapt. Innovation, however, does not wait.
What is emerging instead are parallel propositions: AI companions, early detection tools, radiology and MRI analysis platforms, often built alongside reference doctors, teachers, or schools designing their own solutions.
Nothing appears to change at the policy level. Yet everything shifts quietly, through adoption.
A strategic crossroad for venture capital?
This environment is steadily reshaping how investors look at technology assets, creating a clear bifurcation:
- Legacy SaaS: Traditional SaaS models are thinning out, remaining relevant only if they serve as the essential data layer preceding automation.
- The Fundamentals Track: Sustainable growth focused on "sticky" productivity tools and the low-level stack (infrastructure, networking, core productivity).
- The Agentic Track: The emergence of "Service-as-a-Software" (agent-led) firms showing hyper-growth, though their valuation frameworks remain complex and difficult to pin down
Macro conditions remain destabilized. Geopolitics, regulation, elections, tightening corporate budgets, and constrained liquidity keep valuation spreads volatile. The result feels familiar: long periods of inertia, punctuated by sharp acquisition and M&A waves.
Once again, we believe 2026 will reward systems that absorb shocks, execute under constraint, and compound quietly.
A few weeks into 2026, clearer patterns are beginning to emerge across the tech landscape. Not because uncertainty has faded, but because the system’s pressure points are now coming into view.
Get a glimpse into how Julien-David Nitlech, Managing Partner at IRIS, sees the year likely to unfold. His views also feature alongside peers in The French Tech Journal - French VC Outlook 2026: More concentration, fewer bets, bigger stakes with analysis by Helen O’Reilly-Durand.
Weirdest year ever... until 2027?
2026 opens with a strange sense of déjà vu. The cadence looks familiar: AI mega-fundraises, a race for exposure to AI or defense, capital clustering around a narrow set of narratives. But beneath the surface, the ground is shifting.
There's two structural AI forces are at work.
- First, AI leadership is being internalized. Google, via Gemini, has absorbed the full stack. Others remain structurally dependent on external capital and partnerships that once felt permanent, until they weren’t.
- Second, Chinese AI is driving the marginal cost of “AI as a feature” toward zero, much like cloud did between 2008 and 2010.
The implication is uncomfortable. A large share of current AI valuations assumes linear progress and durable pricing power. A collective realization that “it is complicated” would not be marginal. It would ripple through ongoing and future rounds, with potentially violent effects.
Industrial AI and the infrastructure moat
The AI bull run continues, but its nature is changing. Multimodal and agentic systems are expanding their operational footprint across enterprises. Upstream, a clear premium is emerging around data readiness. Tools that flatten, orchestrate, and industrialize data flows are becoming mission-critical. This is driving valuation tailwinds for data centers, data lakes, and the broader data infrastructure stack.
Hyperscalers continue to consolidate aggressively, using M&A to assemble end-to-end, AI-driven customer experiences rather than isolated features. At the growth stage, companies with deep product penetration are widening their moats by embedding AI directly into existing client bases, particularly in health, manufacturing, and industrial technology.
When AI cheapens, value moves elsewhere
As AI becomes increasingly abundant, its value will relocate.It will move toward execution and toward companies that get things done. Robotics players and industrial production companies operating under intense margin pressure are likely to drive the next wave of adoption.
Software, robotics, and applied AI embedded directly into real operations will matter far more than standalone intelligence. Amazon has already shown us the blueprint. 2026 will surface early signals with real volume likely to be seen in the following years.
After defense, space becomes strategic for Europe
Defense has repriced sovereignty and security, but differentiation and exits are increasingly hard to read. Drone systems and AI-native defense platforms mature quickly, often faster than capital markets can price them. The next strategic layer is space.
Owning payloads, private access to data, connectivity, and imagery will matter in the same way owning cloud or proprietary data once did. Space is shifting from state-led infrastructure to early commercial trade. The first true platforms are now emerging.
Social systems will not reform. They will be bypassed.
Healthcare, education, and retirement systems are structurally unaffordable, and governments are unable to move fast enough to adapt. Innovation, however, does not wait.
What is emerging instead are parallel propositions: AI companions, early detection tools, radiology and MRI analysis platforms, often built alongside reference doctors, teachers, or schools designing their own solutions.
Nothing appears to change at the policy level. Yet everything shifts quietly, through adoption.
A strategic crossroad for venture capital?
This environment is steadily reshaping how investors look at technology assets, creating a clear bifurcation:
- Legacy SaaS: Traditional SaaS models are thinning out, remaining relevant only if they serve as the essential data layer preceding automation.
- The Fundamentals Track: Sustainable growth focused on "sticky" productivity tools and the low-level stack (infrastructure, networking, core productivity).
- The Agentic Track: The emergence of "Service-as-a-Software" (agent-led) firms showing hyper-growth, though their valuation frameworks remain complex and difficult to pin down
Macro conditions remain destabilized. Geopolitics, regulation, elections, tightening corporate budgets, and constrained liquidity keep valuation spreads volatile. The result feels familiar: long periods of inertia, punctuated by sharp acquisition and M&A waves.
Once again, we believe 2026 will reward systems that absorb shocks, execute under constraint, and compound quietly.
A few weeks into 2026, clearer patterns are beginning to emerge across the tech landscape. Not because uncertainty has faded, but because the system’s pressure points are now coming into view.
Get a glimpse into how Julien-David Nitlech, Managing Partner at IRIS, sees the year likely to unfold. His views also feature alongside peers in The French Tech Journal - French VC Outlook 2026: More concentration, fewer bets, bigger stakes with analysis by Helen O’Reilly-Durand.
Weirdest year ever... until 2027?
2026 opens with a strange sense of déjà vu. The cadence looks familiar: AI mega-fundraises, a race for exposure to AI or defense, capital clustering around a narrow set of narratives. But beneath the surface, the ground is shifting.
There's two structural AI forces are at work.
- First, AI leadership is being internalized. Google, via Gemini, has absorbed the full stack. Others remain structurally dependent on external capital and partnerships that once felt permanent, until they weren’t.
- Second, Chinese AI is driving the marginal cost of “AI as a feature” toward zero, much like cloud did between 2008 and 2010.
The implication is uncomfortable. A large share of current AI valuations assumes linear progress and durable pricing power. A collective realization that “it is complicated” would not be marginal. It would ripple through ongoing and future rounds, with potentially violent effects.
Industrial AI and the infrastructure moat
The AI bull run continues, but its nature is changing. Multimodal and agentic systems are expanding their operational footprint across enterprises. Upstream, a clear premium is emerging around data readiness. Tools that flatten, orchestrate, and industrialize data flows are becoming mission-critical. This is driving valuation tailwinds for data centers, data lakes, and the broader data infrastructure stack.
Hyperscalers continue to consolidate aggressively, using M&A to assemble end-to-end, AI-driven customer experiences rather than isolated features. At the growth stage, companies with deep product penetration are widening their moats by embedding AI directly into existing client bases, particularly in health, manufacturing, and industrial technology.
When AI cheapens, value moves elsewhere
As AI becomes increasingly abundant, its value will relocate.It will move toward execution and toward companies that get things done. Robotics players and industrial production companies operating under intense margin pressure are likely to drive the next wave of adoption.
Software, robotics, and applied AI embedded directly into real operations will matter far more than standalone intelligence. Amazon has already shown us the blueprint. 2026 will surface early signals with real volume likely to be seen in the following years.
After defense, space becomes strategic for Europe
Defense has repriced sovereignty and security, but differentiation and exits are increasingly hard to read. Drone systems and AI-native defense platforms mature quickly, often faster than capital markets can price them. The next strategic layer is space.
Owning payloads, private access to data, connectivity, and imagery will matter in the same way owning cloud or proprietary data once did. Space is shifting from state-led infrastructure to early commercial trade. The first true platforms are now emerging.
Social systems will not reform. They will be bypassed.
Healthcare, education, and retirement systems are structurally unaffordable, and governments are unable to move fast enough to adapt. Innovation, however, does not wait.
What is emerging instead are parallel propositions: AI companions, early detection tools, radiology and MRI analysis platforms, often built alongside reference doctors, teachers, or schools designing their own solutions.
Nothing appears to change at the policy level. Yet everything shifts quietly, through adoption.
A strategic crossroad for venture capital?
This environment is steadily reshaping how investors look at technology assets, creating a clear bifurcation:
- Legacy SaaS: Traditional SaaS models are thinning out, remaining relevant only if they serve as the essential data layer preceding automation.
- The Fundamentals Track: Sustainable growth focused on "sticky" productivity tools and the low-level stack (infrastructure, networking, core productivity).
- The Agentic Track: The emergence of "Service-as-a-Software" (agent-led) firms showing hyper-growth, though their valuation frameworks remain complex and difficult to pin down
Macro conditions remain destabilized. Geopolitics, regulation, elections, tightening corporate budgets, and constrained liquidity keep valuation spreads volatile. The result feels familiar: long periods of inertia, punctuated by sharp acquisition and M&A waves.
Once again, we believe 2026 will reward systems that absorb shocks, execute under constraint, and compound quietly.
What’s next: IRIS outlook for 2026
“If we want things to stay as they are, things will have to change.” Archibald Colquhoun
A few weeks into 2026, clearer patterns are beginning to emerge across the tech landscape. Not because uncertainty has faded, but because the system’s pressure points are now coming into view.
Get a glimpse into how Julien-David Nitlech, Managing Partner at IRIS, sees the year likely to unfold. His views also feature alongside peers in The French Tech Journal - French VC Outlook 2026: More concentration, fewer bets, bigger stakes with analysis by Helen O’Reilly-Durand.
Weirdest year ever... until 2027?
2026 opens with a strange sense of déjà vu. The cadence looks familiar: AI mega-fundraises, a race for exposure to AI or defense, capital clustering around a narrow set of narratives. But beneath the surface, the ground is shifting.
There's two structural AI forces are at work.
- First, AI leadership is being internalized. Google, via Gemini, has absorbed the full stack. Others remain structurally dependent on external capital and partnerships that once felt permanent, until they weren’t.
- Second, Chinese AI is driving the marginal cost of “AI as a feature” toward zero, much like cloud did between 2008 and 2010.
The implication is uncomfortable. A large share of current AI valuations assumes linear progress and durable pricing power. A collective realization that “it is complicated” would not be marginal. It would ripple through ongoing and future rounds, with potentially violent effects.
Industrial AI and the infrastructure moat
The AI bull run continues, but its nature is changing. Multimodal and agentic systems are expanding their operational footprint across enterprises. Upstream, a clear premium is emerging around data readiness. Tools that flatten, orchestrate, and industrialize data flows are becoming mission-critical. This is driving valuation tailwinds for data centers, data lakes, and the broader data infrastructure stack.
Hyperscalers continue to consolidate aggressively, using M&A to assemble end-to-end, AI-driven customer experiences rather than isolated features. At the growth stage, companies with deep product penetration are widening their moats by embedding AI directly into existing client bases, particularly in health, manufacturing, and industrial technology.
When AI cheapens, value moves elsewhere
As AI becomes increasingly abundant, its value will relocate.It will move toward execution and toward companies that get things done. Robotics players and industrial production companies operating under intense margin pressure are likely to drive the next wave of adoption.
Software, robotics, and applied AI embedded directly into real operations will matter far more than standalone intelligence. Amazon has already shown us the blueprint. 2026 will surface early signals with real volume likely to be seen in the following years.
After defense, space becomes strategic for Europe
Defense has repriced sovereignty and security, but differentiation and exits are increasingly hard to read. Drone systems and AI-native defense platforms mature quickly, often faster than capital markets can price them. The next strategic layer is space.
Owning payloads, private access to data, connectivity, and imagery will matter in the same way owning cloud or proprietary data once did. Space is shifting from state-led infrastructure to early commercial trade. The first true platforms are now emerging.
Social systems will not reform. They will be bypassed.
Healthcare, education, and retirement systems are structurally unaffordable, and governments are unable to move fast enough to adapt. Innovation, however, does not wait.
What is emerging instead are parallel propositions: AI companions, early detection tools, radiology and MRI analysis platforms, often built alongside reference doctors, teachers, or schools designing their own solutions.
Nothing appears to change at the policy level. Yet everything shifts quietly, through adoption.
A strategic crossroad for venture capital?
This environment is steadily reshaping how investors look at technology assets, creating a clear bifurcation:
- Legacy SaaS: Traditional SaaS models are thinning out, remaining relevant only if they serve as the essential data layer preceding automation.
- The Fundamentals Track: Sustainable growth focused on "sticky" productivity tools and the low-level stack (infrastructure, networking, core productivity).
- The Agentic Track: The emergence of "Service-as-a-Software" (agent-led) firms showing hyper-growth, though their valuation frameworks remain complex and difficult to pin down
Macro conditions remain destabilized. Geopolitics, regulation, elections, tightening corporate budgets, and constrained liquidity keep valuation spreads volatile. The result feels familiar: long periods of inertia, punctuated by sharp acquisition and M&A waves.
Once again, we believe 2026 will reward systems that absorb shocks, execute under constraint, and compound quietly.
A few weeks into 2026, clearer patterns are beginning to emerge across the tech landscape. Not because uncertainty has faded, but because the system’s pressure points are now coming into view.
Get a glimpse into how Julien-David Nitlech, Managing Partner at IRIS, sees the year likely to unfold. His views also feature alongside peers in The French Tech Journal - French VC Outlook 2026: More concentration, fewer bets, bigger stakes with analysis by Helen O’Reilly-Durand.
Weirdest year ever... until 2027?
2026 opens with a strange sense of déjà vu. The cadence looks familiar: AI mega-fundraises, a race for exposure to AI or defense, capital clustering around a narrow set of narratives. But beneath the surface, the ground is shifting.
There's two structural AI forces are at work.
- First, AI leadership is being internalized. Google, via Gemini, has absorbed the full stack. Others remain structurally dependent on external capital and partnerships that once felt permanent, until they weren’t.
- Second, Chinese AI is driving the marginal cost of “AI as a feature” toward zero, much like cloud did between 2008 and 2010.
The implication is uncomfortable. A large share of current AI valuations assumes linear progress and durable pricing power. A collective realization that “it is complicated” would not be marginal. It would ripple through ongoing and future rounds, with potentially violent effects.
Industrial AI and the infrastructure moat
The AI bull run continues, but its nature is changing. Multimodal and agentic systems are expanding their operational footprint across enterprises. Upstream, a clear premium is emerging around data readiness. Tools that flatten, orchestrate, and industrialize data flows are becoming mission-critical. This is driving valuation tailwinds for data centers, data lakes, and the broader data infrastructure stack.
Hyperscalers continue to consolidate aggressively, using M&A to assemble end-to-end, AI-driven customer experiences rather than isolated features. At the growth stage, companies with deep product penetration are widening their moats by embedding AI directly into existing client bases, particularly in health, manufacturing, and industrial technology.
When AI cheapens, value moves elsewhere
As AI becomes increasingly abundant, its value will relocate.It will move toward execution and toward companies that get things done. Robotics players and industrial production companies operating under intense margin pressure are likely to drive the next wave of adoption.
Software, robotics, and applied AI embedded directly into real operations will matter far more than standalone intelligence. Amazon has already shown us the blueprint. 2026 will surface early signals with real volume likely to be seen in the following years.
After defense, space becomes strategic for Europe
Defense has repriced sovereignty and security, but differentiation and exits are increasingly hard to read. Drone systems and AI-native defense platforms mature quickly, often faster than capital markets can price them. The next strategic layer is space.
Owning payloads, private access to data, connectivity, and imagery will matter in the same way owning cloud or proprietary data once did. Space is shifting from state-led infrastructure to early commercial trade. The first true platforms are now emerging.
Social systems will not reform. They will be bypassed.
Healthcare, education, and retirement systems are structurally unaffordable, and governments are unable to move fast enough to adapt. Innovation, however, does not wait.
What is emerging instead are parallel propositions: AI companions, early detection tools, radiology and MRI analysis platforms, often built alongside reference doctors, teachers, or schools designing their own solutions.
Nothing appears to change at the policy level. Yet everything shifts quietly, through adoption.
A strategic crossroad for venture capital?
This environment is steadily reshaping how investors look at technology assets, creating a clear bifurcation:
- Legacy SaaS: Traditional SaaS models are thinning out, remaining relevant only if they serve as the essential data layer preceding automation.
- The Fundamentals Track: Sustainable growth focused on "sticky" productivity tools and the low-level stack (infrastructure, networking, core productivity).
- The Agentic Track: The emergence of "Service-as-a-Software" (agent-led) firms showing hyper-growth, though their valuation frameworks remain complex and difficult to pin down
Macro conditions remain destabilized. Geopolitics, regulation, elections, tightening corporate budgets, and constrained liquidity keep valuation spreads volatile. The result feels familiar: long periods of inertia, punctuated by sharp acquisition and M&A waves.
Once again, we believe 2026 will reward systems that absorb shocks, execute under constraint, and compound quietly.
A few weeks into 2026, clearer patterns are beginning to emerge across the tech landscape. Not because uncertainty has faded, but because the system’s pressure points are now coming into view.
Get a glimpse into how Julien-David Nitlech, Managing Partner at IRIS, sees the year likely to unfold. His views also feature alongside peers in The French Tech Journal - French VC Outlook 2026: More concentration, fewer bets, bigger stakes with analysis by Helen O’Reilly-Durand.
Weirdest year ever... until 2027?
2026 opens with a strange sense of déjà vu. The cadence looks familiar: AI mega-fundraises, a race for exposure to AI or defense, capital clustering around a narrow set of narratives. But beneath the surface, the ground is shifting.
There's two structural AI forces are at work.
- First, AI leadership is being internalized. Google, via Gemini, has absorbed the full stack. Others remain structurally dependent on external capital and partnerships that once felt permanent, until they weren’t.
- Second, Chinese AI is driving the marginal cost of “AI as a feature” toward zero, much like cloud did between 2008 and 2010.
The implication is uncomfortable. A large share of current AI valuations assumes linear progress and durable pricing power. A collective realization that “it is complicated” would not be marginal. It would ripple through ongoing and future rounds, with potentially violent effects.
Industrial AI and the infrastructure moat
The AI bull run continues, but its nature is changing. Multimodal and agentic systems are expanding their operational footprint across enterprises. Upstream, a clear premium is emerging around data readiness. Tools that flatten, orchestrate, and industrialize data flows are becoming mission-critical. This is driving valuation tailwinds for data centers, data lakes, and the broader data infrastructure stack.
Hyperscalers continue to consolidate aggressively, using M&A to assemble end-to-end, AI-driven customer experiences rather than isolated features. At the growth stage, companies with deep product penetration are widening their moats by embedding AI directly into existing client bases, particularly in health, manufacturing, and industrial technology.
When AI cheapens, value moves elsewhere
As AI becomes increasingly abundant, its value will relocate.It will move toward execution and toward companies that get things done. Robotics players and industrial production companies operating under intense margin pressure are likely to drive the next wave of adoption.
Software, robotics, and applied AI embedded directly into real operations will matter far more than standalone intelligence. Amazon has already shown us the blueprint. 2026 will surface early signals with real volume likely to be seen in the following years.
After defense, space becomes strategic for Europe
Defense has repriced sovereignty and security, but differentiation and exits are increasingly hard to read. Drone systems and AI-native defense platforms mature quickly, often faster than capital markets can price them. The next strategic layer is space.
Owning payloads, private access to data, connectivity, and imagery will matter in the same way owning cloud or proprietary data once did. Space is shifting from state-led infrastructure to early commercial trade. The first true platforms are now emerging.
Social systems will not reform. They will be bypassed.
Healthcare, education, and retirement systems are structurally unaffordable, and governments are unable to move fast enough to adapt. Innovation, however, does not wait.
What is emerging instead are parallel propositions: AI companions, early detection tools, radiology and MRI analysis platforms, often built alongside reference doctors, teachers, or schools designing their own solutions.
Nothing appears to change at the policy level. Yet everything shifts quietly, through adoption.
A strategic crossroad for venture capital?
This environment is steadily reshaping how investors look at technology assets, creating a clear bifurcation:
- Legacy SaaS: Traditional SaaS models are thinning out, remaining relevant only if they serve as the essential data layer preceding automation.
- The Fundamentals Track: Sustainable growth focused on "sticky" productivity tools and the low-level stack (infrastructure, networking, core productivity).
- The Agentic Track: The emergence of "Service-as-a-Software" (agent-led) firms showing hyper-growth, though their valuation frameworks remain complex and difficult to pin down
Macro conditions remain destabilized. Geopolitics, regulation, elections, tightening corporate budgets, and constrained liquidity keep valuation spreads volatile. The result feels familiar: long periods of inertia, punctuated by sharp acquisition and M&A waves.
Once again, we believe 2026 will reward systems that absorb shocks, execute under constraint, and compound quietly.
