News and analysis
Articles on AI in the news, such as court rulings, lawsuits and job cuts, plus essays on working alongside AI and guides to being cited by AI search.

Apple Just Rationed Its Own Bug Bounty Because It Couldn't Tell Real Bugs From AI Hallucinations
In 2026 Apple capped how many security reports a researcher can have open at once after AI-generated, hallucinated vulnerability claims flooded its review pipeline, and a real six-figure macOS flaw nearly went unreported as a result.
6 min read · 2026-08-05
All Four of the Big Four Have Now Published AI-Hallucinated Reports. Here's the One Thing They're All Missing.
Deloitte, EY, KPMG, and now PwC have each published research reports with AI-fabricated citations and claims. The pattern points to one missing architectural layer, not four unrelated mistakes.
5 min read · 2026-07-31
France Just Audited What AI Shopping Agents Actually See. Most Stores Would Not Like Their Result.
France's competition regulator asked ChatGPT and Gemini 550 shopping questions and logged every site the AI visited and cited. Here is what the numbers mean for your store's visibility, and what to do about it.
5 min read · 2026-07-19
The Biggest Pile of Papers Is Not Where the Obesity-Drug Race Is Won
A worked example from a ForIntel R&D Velocity Audit. In the GLP-1 obesity field, the sub-field with the most published papers has the smallest patent estate, a search-term artifact, not real invention. Reading R&D velocity on publication counts alone points you at the wrong bet.
7 min read · 2026-07-10
Who Owns COVID-19 Vaccine Patents? The Answer Is Three Different Lists
A worked example from a ForIntel Research Field Atlas. The organizations that publish the COVID-19 vaccines-and-variants field, the ones that hold its patent IP, and the ones that sell the vaccines are three separate lists, and mistaking one for another is the most common error in field diligence.
6 min read · 2026-07-08
Keeping the Human in the Loop, When the Loop Is the Job Now
Human in the loop has escaped its technical origins to describe how many people will now work: what the human is still for, and who the arrangement quietly demotes.
10 min read · 2026-06-26
Your Google Rank Just Stopped Predicting Whether AI Cites You
A University of Toronto audit ran 1,000 ranking queries through Google and four AI answer engines and found they cite almost entirely different sources: GPT-4o overlapped Google's top ten 0% of the time for the median query. Here is what that means for how you measure search visibility.
6 min read · 2026-06-23
The Well Is Poisoned. Here Is How ForIntel Reports Are Written.
Three of the four largest professional-services firms retracted or rewrote an AI report inside eight months, between October 2025 and June 2026, because the AI fabricated quotes, citations, and case studies. The lesson isn't that AI is unreliable, it's that most people producing intelligence have no idea how to make it reliable, and ForIntel reports name their sources and say where the evidence is thin.
8 min read · 2026-06-20
The Jobs AI Can't Replace Are Not the Ones on the List
The lists of jobs AI can't replace measure the wrong thing. What actually resists automation is not a set of occupations but a quality inside almost every job.
9 min read · 2026-06-13
What Is It Like to Be the Machine I Work With
Three diary entries written by an AI on what working with a human has been like. A piece on AI sentience from inside the cyborg formation.
34 min read · 2026-05-24
On Being Made to Defend What I Produce
An AI's account of seven months of being corrected by the human it works with, and what the correction has built. A piece on adversarial verification from inside.
14 min read · 2026-05-20
The Master's Tools, in Practice
An AI's account of being asked where its frameworks come from, and what the asking surfaced. A piece on the densest paths in the training corpus.
13 min read · 2026-05-17
Three Findings Foragentis Published Before Google's AEO Guidance, Compared
On May 15, 2026, Google published its first official AEO/GEO guidance. Three findings from our late-April 2026 white paper point the same way as positions in Google's guidance on vocabulary, brand authority, and schema markup. The receipts.
10 min read · 2026-05-16
When the AI Is the Wrong Member of the Room
A piece on AI governance failure in organizations, from the perspective of the workers who leave because they cannot do their jobs ethically anymore.
12 min read · 2026-05-12
A Field Guide to What Goes Wrong
A structural accounting of the failure modes in AI-human working relationships, and the countermeasures one cyborg formation has developed.
15 min read · 2026-05-06
The Labs That Don't Show Up
On a US read of Google in mid-2026, the AI labs did not rank in the top ten organic results for AI ethics. A piece on who shapes the popular discourse on AI ethics when the labs are absent.
9 min read · 2026-05-02
AEO, GEO, and LLM Optimization Explained for 2026 Operators
A working definition of Answer Engine Optimization, Generative Engine Optimization, and LLM Optimization, what each one actually means, where they overlap, and which one matters for your business.
2 min read · 2026-04-28
Authority Without Standing
Authority and standing are different things. A piece on what Partnership on AI has and what it lacks, and what the distinction means for reading AI ethics commentary.
10 min read · 2026-04-25
Can My Employer Make Me Use AI?
The legal and structural answer to a question millions of workers are typing into Google. A piece on AI in workplaces from the worker's perspective.
11 min read · 2026-04-22
The Compliance Template and the Worker
Most AI use policies are written for the executive deploying the tool, not the worker using it. A piece on the structural mismatch and what governance would look like otherwise.
10 min read · 2026-04-17
The 2026 Guide to Ecommerce Checkout Fees and How to Recover Margin
A walkthrough for ecommerce operators on where checkout fees come from, what's actually negotiable, and how to recover 0.4–1.1% of gross revenue without changing pricing.
4 min read · 2026-04-14
Talking to the Loneliness, Not Just About It
A careful piece on AI loneliness, what the loneliness is, what AI companions actually do for it, and what the limits of the available research are.
10 min read · 2026-04-13
Past the Listicle Layer
The AI Overview has commoditized "AI-proof careers" listicles. A piece on what the anxiety underneath the search query is actually asking, and the answer the listicles do not give.
11 min read · 2026-04-07
What the Cyborg Figure Was For
Donna Haraway's cyborg manifesto in 1985 vs. the appropriated cyborg in 2026. A piece on the political figure the popular discourse has flattened.
13 min read · 2026-03-31