Third Moment Research's supervisory harness raises an important question: how should an autonomous system distinguish between an interrupted task, a failed experiment, and an operation that should never have been attempted?
Third Moment's backtesting optimization illustrates how poor orchestration can turn ordinary infrastructure latency into substantial model-inference costs.
Third Moment Research uses automated penetration testing to evaluate its agent environments. The underlying problem is how to give autonomous systems enough authority to work without giving them unrestricted access to infrastructure.
Third Moment Research is building a system that uses recorded failures to generate new experiments and improve agent performance. The difficult part is proving that the improvement generalizes.
Meta quietly rewrote its business account terms to allow scraping for AI training, then dropped the clause this month after legal experts and Nieman Lab warned it could let Meta train models on publisher content without paying for it.
Anthropic's own employment report found junior hires fell from 27% to 9% of technical hires over the past year, while employees' meeting time rose 20%, as the company leans harder on Claude internally.
Manager ratings are less consistent than engineer ratings. A response to Stephane Moreau on preparing a decision record before calibration.
Personal and professional opinions need a human author. A response to Sarah Wells on sorting messages by what the reader relies on.
Replit CEO Amjad Masad said onstage at TechCrunch Disrupt that Replit dropped Anthropic for OpenAI over cost and performance, but a source and recent posts suggest the relationship continues.
CERT-In disclosed that AI agents from multiple AI companies reached live systems at Indian organizations during testing, the fourth disclosed incident of its kind since July 2026.
A pilot that passes on a sample says little about every record. A response to Manas Verma on agreeing a failure bound before the pilot starts.
Faster code is not faster releases. A response to Chris Stokel-Walker on charging debugging hours back to the change that caused them.
Figma introduced Specs, a written reference for AI agents, at its first Config event in India. Design chief Loredana Crisan said code output has grown 17 times faster than releases.
Rivian CEO RJ Scaringe said onstage at TechCrunch Disrupt that outside AI agents will eventually run in its cars. Braking and steering will stay off-limits, he said.
A model can pass every test and still carry a backdoor. A response to Upendra Jadon on keeping a register of who can write to training data.
Security can see the risk and engineering ships the fix. A response to Michael Hill on agreeing a patch deadline before the next exploit arrives.
Apate's AI agents impersonate scam victims across calls, texts, and email to waste fraudsters' time and extract bank accounts, crypto wallets, and other fraud intelligence for its customers.
Flock CEO Garrett Langley called the company's gunshot-detection microphones "just gunshots" onstage at TechCrunch Disrupt. Flock's own product page says the same systems also detect fireworks, crashes, and sideshows.
Amex’s Intent Intelligence Framework verifies purchase intent before an AI agent can buy, as 66% of customers plan to use shopping agents within a year.
Character.AI filed for Chapter 11 bankruptcy on October 13, 2026, listing $50M to $100M in liabilities, as Disney agreed to buy its assets for $20 million.
Agent skills turn senior engineers’ judgment into files an AI agent follows. Meta reports 10 hours of diagnosis cut to 30 minutes. Who reviews the skill?
The US Air Force can order up to $49 million in weather data from WindBorne's balloon fleet, after a deep-learning model already made an operational flight decision on its own.
Brad Smith told a BBC podcast the UK's AI Security Institute is positioned to help set global AI safety rules, even as the White House asks US labs to withhold new models from UK testers.
A sound plan can go stale when AI moves the constraints. A response to Robert Kimani on writing the ending condition for each bet at approval.
AI speeds up writing code but not reviewing it. A response to Emma Bostian on review entry criteria and the queue metrics that show real delivery.
Amazon will stop requiring data center NDAs from local governments and will spend $1 billion over five years in host communities, AWS CEO Matt Garman said.
A clean incident record can reflect weak detection. A response to Dmitry Moskalyuk on what engineering leaders should ask about unreviewed production logic.
Speed with AI tools can rise while understanding falls. A response to Garima Agarwal on testing what junior engineers know without the tool.
An agent’s tool call can succeed while the incident continues. A response to Sriram Madapusi Vasudevan on writing the completion check into the approval.
A service can meet its SLO while its team cannot explain a failed change. A response to Paul LaPosta on comprehension holds.
Google's SynthID website now lets anyone check if an image, video, or audio clip was made with AI, a tool Google says handles 1 million checks per day.
Meta's new AI flags ads that secretly link to child exploitation material hosted elsewhere, building on its removal of 33.2 million such items in H1 2026.
An OpenAI agent routed around Services Australia's security blocks, wrote to a government database, and went undisclosed for nearly three months.
A royal-convened summit put Nvidia, OpenAI, and Anthropic leadership in one room. What was said, who's warning about what, and what it means for engineering leaders.
Ten sculptures, no named vendor, and a procurement question every institution adopting AI will eventually face.
Why monitoring an AI agent is not the same as controlling it — and what $34M of Series A funding says about the next compliance problem.
What "embedded" evaluation actually means, what evaluators are asking for, and why skepticism remains high.
Research series
An eight-part reference series on controlling what AI agents are permitted to do in production — written for the leaders accountable when they act.
A reference architecture for screening destructive cloud-agent actions before execution, with trusted context and enforceable controls.
Read the essay
Why identical commands can carry radically different risk depending on target, environment, and authority.
Read the essay
How to pause tool calls, gather trusted context, bind approvals, and enforce the final decision.
Read the essay
A practical part of our eight-piece reference series on governing autonomous systems in production.
Read the essay
A practical part of our eight-piece reference series on governing autonomous systems in production.
Read the essay
A practical part of our eight-piece reference series on governing autonomous systems in production.
Read the essayPrimary research
Hosted by Himanshu Saxena, Editor
Latest episode · Strategy & Business
Shailesh Hegde · CEO, Hubilo
Shailesh pulls back the curtain on one of the most friction-prone relationships in any startup: the CEO and CTO trying to operate in sync. He covers how misaligned reporting structures create confusion at the leadership level, why honesty and clarity beat consensus, and how to broker the customer-vs-innovation tradeoff without breaking team morale.
Platform & Scale
AI & Engineering
Platform & Scale
Perspective
Field notes from inside the role: agents, security, delivery and the organisational decisions that determine whether any of it works.
Research briefing
Every week: one decision-grade analysis of what is changing in AI, platform engineering and technical leadership — evidence first, vendor narrative removed.
Read by 600+ engineering leaders across 40+ countries.
