Market Pulse
Market Pulse
Every Acquisition Is a Theory
Dynatrace paid $915 million for Arize and classified agents as applications to be traced and scored. Fortinet's purchase of Virtue AI classified them as attack surface. Okta buying Permiso made them non-human identities that need governing after they authenticate. Each deal makes sense on its own. Read together, they're doing something less obvious: every acquisition bakes in a theory about what an agent is, and that theory decides which problems the product will catch for you and which ones it won't. Choosing agent infrastructure right now means choosing a frame, whether or not you meant to.
Every Acquisition Is a Theory
Dynatrace paid $915 million for Arize and classified agents as applications to be traced and scored. Fortinet's purchase of Virtue AI classified them as attack surface. Okta buying Permiso made them non-human identities that need governing after they authenticate. Each deal makes sense on its own. Read together, they're doing something less obvious: every acquisition bakes in a theory about what an agent is, and that theory decides which problems the product will catch for you and which ones it won't. Choosing agent infrastructure right now means choosing a frame, whether or not you meant to.
Research Briefing
Agent-Native Telemetry: Verifiable State-Delta Evidence for Autonomous Operations
The paper treats telemetry as simultaneously an observability, security, and governance artifact — a combination no single vendor category currently owns.
Tamper detection is demonstrated, but proving that nothing important was omitted from the record remains only partially addressed.
Research Briefing
Long-Horizon Agent Trajectory Attribution: A Unified Benchmark and Fine-Grained Annotation Framework
Outcome benchmarks grade results; attribution identifies which specific moment inside a trajectory seeded a failure or unsafe action.
Assigning responsibility within trajectories is prerequisite for auditable records, but reliability in contested real-world disputes remains untested.
Research Briefing
ScrambleToolBench: Agents Search Exhaustively Even When Their Own Map Points to the Next Step
Successful calls under outdated mappings produce no visible errors, making schema drift harder to catch than outright tool failures.
If a tool's actual behavior shifts but the agent's belief doesn't update, prior permissions may no longer cover what's really happening.
Research Briefing
Recursive Synthesis for Long-Horizon Terminal Tasks
Dropping the cost of verified training tasks by orders of magnitude could reshape how agent capabilities get benchmarked at scale.
Synthetic tasks may not capture production failure modes involving changing permissions, authority boundaries, and real-world exceptions.
Price Signal
Google launched Gemini 3.7 Flash on August 13 at $0.75 per million input tokens and $3.75 per million output tokens, with both rates doubling on January 1, 2027. The model is positioned as Google's "workhorse for coding and agents" and already powers its own agent infrastructure by default.
The promotional window looks like a discount. It functions as something closer to a subsidy for adoption during the period when architectural choices get made. Agent workflows involve calibrating loop lengths, retry budgets, evaluation frequency, and the scope of what gets automated against a cost assumption. Once those calibrations are set and running in production, they tend to stay. Developers don't casually redesign a working pipeline because the price of inference changed.
Google reinforced this by applying the same introductory rate to 3.6 Flash, eliminating any price difference between model generations during the window. You migrate to 3.7 because it's more capable, not because it's cheaper. The capability preference becomes the production default. The production default carries the doubled rate into 2027.
The question practitioners should be asking is whether cheaper inference during this window actually reduces their total cost, or whether it funds more ambitious automation that meets the permanent rate with greater exposure. Promotional pricing shapes what people build. What people build determines what they're committed to when the promotion ends.
Reading List
Past Articles
Eighteen percent of firms formally use AI, according to the Census Bureau — a number now doing heavy work in strategy de...
Watch a coding agent work for an hour. Forty files, tests written and failed and rewritten, and when it stops, no user h...
Somewhere in a traffic policy sits the suggestion that you break your automation into three crawlers, one per purpose, s...
Visa's rules now define agentic transactions and hold cardholders responsible for what their agents do. Real agents are ...