Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsWhen an AI agent acts between a company and its customer, who keeps the customer relationship—and who may use the data the agent creates? That question recurred at Madrona’s 2026 IA40 Summit, but the event produced no settled answer. Its other takeaways pointed to a broader shift: agents could change how people use software and shop, while human processes, security and competing views on provider dependence shape how quickly businesses can deploy them.
What was the Madrona IA40 Summit?
Madrona’s annual gathering of AI startups, investors and technology executives took place Sept. 29–30, 2026, at the Four Seasons in Seattle. The official agenda included agentic data, AI harnesses, pilots and return on investment, collaborative agents, trust, web tools, software and autonomous systems. GeekWire’s event report and the official IA40 summit page provide the event context.
As an Amazon Associate I earn from qualifying purchases.
Across those topics, a practical business question kept surfacing: if an agent mediates an interaction, who retains access to the customer and to the resulting record? Madrona Managing Director Matt McIlwain said the issue of who gets to use data generated through people’s engagement with AI systems came up “over and over and over again.”
Recommended Free Tools
Who controls the customer relationship and agent data?
The summit did not resolve who owns an agent’s work record, its mistakes and corrections, or the data generated during a customer interaction. Moderator Raphaëlle d’Ornano asked whether that record belongs to the customer; GeekWire reported that she had never received a clear answer. Anthropic CTO Rahul Patil did not answer the ownership question directly. He said providers would use available data to improve agent systems.
#1 Best Overall
That comment is not evidence of a contractual right to customer data. The event report offers no legal analysis or contract terms, so it cannot establish who owns a particular record or what a provider may do with it. Those questions depend on the relevant agreements and circumstances, not on a general answer from the summit.
Could agents become the main way people use software?
Some speakers predicted that agents will increasingly act as an interface between users and business applications. Microsoft’s Charles Lamanna said most business software will eventually be used by AI assistants on a user’s behalf. Town CEO Jean-Denis Greze described assistants operating a browser or computer through its interface, even without an API, and predicted a shift toward “thin apps.”
“Everything is going to become a thin app because the better the AI gets at using the computer, the less the app matters as a unit of software,” said Greze.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.
These were speakers’ forecasts, not established outcomes. If agents do become a common software interface, the application may matter less to the end user while control over the agent, its data access and the customer interaction becomes more consequential.
What changes when an agent shops for a customer?
Agent-mediated shopping raises a tension between convenience for the buyer and control of the sale for the merchant. Stripe’s Maia Josebachvili said agent commerce on Stripe had been roughly flat for eight or nine months before rising sharply over the six weeks before the summit. That was her observation about Stripe activity, not a market-wide statistic.
Josebachvili argued that merchants can lose opportunities for add-on sales and advertising when agents make purchases on a customer’s behalf. GeekWire also reported that Amazon had blocked Meta’s Muse assistant from shopping on its site the previous month. Together, those examples illustrate why a merchant may care about whether an agent can access its site and how much of the shopping relationship it retains.
Why is enterprise deployment still difficult?
Speakers described organizational readiness and operational controls—not just model capability—as barriers to putting agents to work. Goldman Sachs’ Archana Vemulapalli put it succinctly: “The bottleneck is actually not AI. The bottleneck is human.” She was referring to roles and processes designed before AI, which may need to change for agents to fit into real work.
AWS’s Swami Sivasubramanian said teams building agents still needed to solve security, identity and monitoring before rollout. He also described pairing generative models with separate systems that check outputs against company rules. Carnegie Mellon professor Zico Kolter argued that system control must keep pace with AI capability, potentially requiring slower development.
For a company assessing an agent platform, the summit’s concerns suggest practical questions to examine:
- Data and context: What information can the agent access, and what happens to the interaction and work records it generates?
- Identity and security: How are users and agents identified, and what permissions limit their actions?
- Monitoring and auditability: Can the organization see what an agent did, including errors and corrections?
- Interoperability: Can the system work with existing applications and tools, including through browser or computer interfaces?
- Rules and outcomes: How are outputs checked against company policies, and does the system deliver measurable work or business value?
The summit reporting raises these evaluation dimensions but does not compare products or rank vendors.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Should companies commit to one AI provider or preserve choice?
Executives disagreed on how much dependence on a model provider is acceptable. Patil argued that keeping the option to switch providers can push companies to build around the least common denominator, diverting engineering effort from work that differentiates the business. Noeri co-CEO Carlos Guestrin argued that intelligence should not be controlled by one or two model companies and that businesses should be able to build and own AI systems. Factory’s Eno Reyes described companies that see no path forward without ceding control to one AI lab, while former GitHub CEO Thomas Dohmke emphasized developer choice.
These are competing strategic positions, not a single summit recommendation. The trade-off is between concentrating on a provider’s capabilities and preserving room to choose or build alternatives.
What do Madrona’s IA40 figures say about the market?
Madrona’s figures for its 2026 IA40 list show how concentrated funding was within that particular cohort. In its article, “Introducing the 2026 Intelligent Applications 40”, Madrona says the 45 companies on the 2026 list had raised $410 billion since founding. Anthropic, OpenAI and Databricks accounted for $377 billion, or 92% of the cohort total. Madrona categorized the funding data as of Aug. 15, 2026.
These are Madrona’s figures for its IA40 companies, not a count of funding across the whole AI sector. Madrona also reported that 23 of the 40 prior-year winners returned to the 2026 list, a 58% repeat-winner rate. Its article says security and privacy rank among the top three purchase criteria for 78% of surveyed enterprises, based on Madrona’s proprietary enterprise survey. It also reports that 52% of AI deals close in under six months and that 41% of enterprises say engineering teams discover AI tools through testing; the article excerpt does not give the underlying survey method or field dates for those two figures.
What the summit left open
The event’s clearest unresolved issue was not whether agents can perform tasks, but who controls the relationship and records when they do. Agent interfaces may alter how customers reach businesses and how businesses reach buyers, yet the event did not settle who may use the resulting data. For companies, the reported concerns point to a linked set of decisions: redesign work around agents, build security and monitoring into deployment, and choose how much control to retain over providers, systems and customer access.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
