Insights from OBSOLETE: 8 AI shifts reshaping your business
AI is breaking the assumptions your business model, workflows, and organization were built on. Intelligence is becoming abundant. Traditional business models are breaking.
AUTONOMOUS: OBSOLETE brought senior leaders from Glean, Mars, Logitech and more to ask the uncomfortable questions: where value is being created, where it’s eroding, and what your organization needs to become to not be obsolete.
Here’s what surfaced across the sessions.
Didn’t catch the event live? Register here to access the sessions on demand
1. Three forces are dismantling the old model at once
The disruption is converging from three directions simultaneously.
- Direct substitution by AI-native entrants. Software companies charging $50-100k+ are being replaced by internal AI agents at a fraction of the cost. Some called it the SaaS apocalypse.
- Client self-sufficiency. Mid and junior employees now produce 30-50% of previous vendor output in-house. Clients are questioning contracts and renegotiating from later starting points. The relationship shift looks permanent.
- The erosion of opacity. The trust premium is disappearing. AI makes the black box transparent. Clients are developing real intuitions about the work and the cost behind it. Pricing structures built on opacity don’t survive that.
2. When intelligence gets cheap, scarcity relocates
It doesn’t disappear. It moves. The scarce thing becomes whatever can’t be separated from the product and handed to an agent.
The premium shifts toward presence, provenance, accountability, trust, judgment, and IP. The behavioral evidence is already visible in how the ultra-wealthy spend:
- Live performances over infinite media options.
- Small venues and personal relationships over scale.
- Human connection over the cheaper technological alternative.
When something is everywhere, the thing that can’t be replicated commands the price.
3. Pricing is being unbundled
Traditional pricing bundled everything together: intelligence, judgment, relationships, access, and IP. One number for all of it. AI is pulling those elements apart and pricing them individually.
- The intelligence portion reprices sharply downward.
- The judgment layer is often worth more than anyone previously credited.
- The winners will know exactly which they’re selling. The losers will keep discounting the wrong one.
Join us at the world’s largest summit for AI innovators.
September 9-10, 2026
+15,000 attendees
Virtual summit
4. New service models are now viable at scale
Cheap intelligence opens models that were never economically possible before. Examples:
- Healthcare: continuous, personalized, always-on support.
- Financial services: real-time guidance that adapts to how each person spends and saves.
The opportunity is to build what couldn’t exist before. Defending what already exists leaves that ground to someone else.
5. The hardest barrier is internal
Organizations were designed around intelligence scarcity. Every process, every approval chain, every bottleneck assumes expensive analysis and limited human capacity. AI replaces those tasks. Then it makes the entire structure look unnecessary.
The problem is who holds the power:
- Partners whose business books sit in the intelligence layer.
- Product teams whose workflows are about to disappear.
- Leadership structures built around functions that AI is quietly absorbing.
The people with the most stake in the current model are usually the ones with the most power to block the change. The technology is ready. The organizations are not.
6. Shared context beats scattered copilots
Most enterprises are deploying isolated copilots. Each one is smart on its own and blind to everything else. That fragmentation creates coordination overhead instead of removing it.
- Agents should work from the same permission-aware source of truth.
- Teams should organize around outcomes, not data location.
- Strategy should stay vendor-neutral, with durable data access and procurement discipline around portability.
Shared enterprise context is what turns scattered tools into a system.
7. Fin proves the rebuild works
Formerly Intercom, the company was in decline. It pivoted hard into AI, renamed itself FIN, and rebuilt around an agent-first engineering model.
- A 2x productivity target hit in nine months.
- Backlog zero approaching across multiple teams.
- Quality improving week over week, even through AI-heavy change.
- Delivering millions of support and sales interactions, and outperforming projections while public SaaS stocks slide.
That’s what it looks like to stop defending the old model and rebuild around the new one.
8. Every industry, same pressure
- Banking (Bunq): 80% automated. 750 people doing the work of 2,000 to 4,000.
- Pharma (GSK): R&D heavily AI-driven. The commercial side is proving harder.
- Telecom (KPN): Network advantage eroding. Betting on service-layer distinctiveness.
- Consumer goods: Mass scaling giving way to one-on-one personalization.
Every one of them is also fighting the same arms race on fraud and security. AI enables your business. It also enables the attack.
The takeaway
Defending the current model is not a strategy. Intelligence reprices down. Judgment reprices up. New models open that were impossible a year ago.
The companies that win won’t be the ones running better AI than their rivals. They’ll be the ones willing to become a different kind of company.
OBSOLETE diagnosed what breaks. AUTONOMOUS 2026 is where we build what comes next. Two days, September 9-10, fully virtual. We bring leaders together to close that transformation gap, and discover what it takes to build an AI-native organization.
Join us at autonomoussummit.ai and figure out what your organization needs to become to make AI pay off.
