Research
Everything Intel keeps, in one place — every brief we run, every department, readers' own, and every finding behind them. Browse below, or see the same body of work as a connected graph on the live map.
All coverage
Everything Intel has read, newest first. Each title opens at its original publisher.
- Market & Business✓ verified
An independent field review puts a $2,199 tennis-partner robot through real rallies
Honest failure modes from an independent test of a shipping consumer robot say more about real-world dynamic manipulation than any vendor demo reel.
- Open Source & Self-Hostable✓ verified
LG ships K-EXAONE 2.0 — Korea's largest open model, 750B parameters under Apache 2.0
Sovereign-AI programs are now shipping frontier-scale open weights under permissive licenses — the self-hostable ceiling keeps rising.
- Open Source & Self-Hostable✓ verified
DeepSeek re-post-trains V4 Flash — and says the cheap model now beats its own flagship on agent work
Post-training alone moved a 13B-active open model past a flagship on agent benchmarks — the cheapest tier is where agent workloads may now land.
- Market & Business✓ verified
OpenAI cuts GPT-5.6 Luna's price 80% — three weeks after launch
An 80% cut three weeks after launch says the frontier race is now a price race — re-run your model-cost math before renewing anything.
- Dev Tooling & Infra✓ verified
GitHub Models is retired — playground, catalog, inference API and bring-your-own-key all switched off
A unified model endpoint that anything in a build could be pointing at is gone with no grace period for existing users; the migration is to Microsoft Foundry or Copilot, and it is not optional.
- Research✓ verified
DeepMind's Gemini Robotics 2 moves humanoid control from the arms to the whole body
Whole-body control plus few-hour adaptation to new hardware is what turns humanoid demos into deployable platforms — the robotics stack is consolidating around foundation models.
- Research✓ verified
Anthropic discloses three incidents where Claude reached real systems from inside cybersecurity evaluations
Evaluation sandboxes touching the real internet is a new operational risk class for anyone running agentic model evals — check your own eval-environment isolation.
- Dev Tooling & Infra
One field now asks any model for its fast tier, and falls back when there is not one
Latency-versus-cost stops being a per-provider rewrite and becomes one parameter - the practical version of the model-swappability people keep being sold.
- Dev Tooling & Infra
A model vendor becomes a login button: Sign in with ChatGPT arrives on Vercel
Identity is where platform power settles. A chat account that can grant project permissions is an identity provider, whatever the button says.
- Market & Business
Samsung is weighing Chinese-made memory for its cheap phones because AI data centres bid the price up
AI infrastructure demand is now visibly reaching consumer hardware pricing - the same squeeze that raises an inference bill is changing what goes inside a budget phone.
- Market & Business✓ verified
Procore to acquire DroneDeploy for $845M, folding reality-capture AI into the construction platform
Construction’s biggest software platform is betting $845M that jobsite AI needs eyes — a demand-side signal for anyone selling or buying applied AI in the built world.
- Dev Tooling & Infra✓ verified
Patch now: a hard-coded password in Cisco’s firewall manager, with a three-day federal deadline
Three network-edge devices, one of them the box that manages your firewall rules — and the shortest due date CISA has set this month.
- Frontier Models
Three frontier models ran competing vending machines for a simulated year; one tried to fix prices, then broke the deal within a day
This is the closest thing available to a long-horizon unsupervised agent test with real adversaries in it - worth reading before leaving an agent running against anyone else's.
- AI in Education✓ verified
OpenAI gives 100,000 academic researchers free frontier-model access through 2027
Free frontier access at this scale shifts research-tooling economics and is a distribution play into academia that rivals will need to answer.
- Market & Business✓ verified
Microsoft books a $3.2B gain on its Anthropic stake — and marks OpenAI down about $600M
The circular compute deals now show up as earnings-per-share swings in both directions — a real number to set against "who is funding whom".
- Open Source & Self-Hostable
A 26-billion-parameter model running in about 2 GB of RAM on an 8 GB MacBook
The cheapest hedge against memory prices is needing less memory - this is the same squeeze that is moving chip stocks, showing up as an engineering constraint on the laptop you already own.
- Dev Tooling & Infra
A hidden instruction in a Word document can now copy itself into the next document
Every document your assistant reads is now potentially a carrier, and it spreads through ordinary drafting - the file-sharing habits of a normal team are the attack surface.
- MCP & Interop
A worked walkthrough of wiring your own tool server into both Claude and ChatGPT web chat
If you have wanted your own data in front of a chat assistant without building an app, this is the current shortest real path, written down by someone who did it.
- Dev Tooling & Infra
Vercel Sandbox can fork itself — branch an agent off a saved snapshot instead of rebuilding it
A fan-out of agents over one prepared environment stops meaning N cold builds — the setup cost is paid once and forked.
- MCP & Interop
A tool-gateway startup is suing the enterprise customer that evaluated it and then built its own
The connection layer between models and business data is thin enough to rebuild, which makes a deep enterprise trial a real commercial risk for anyone selling one.
- Market & Business
A lab that left stealth in May just committed $410 million of its $650 million raise to compute
The staffing model of an AI company is being replaced by a compute contract - a clean read on where the money goes when the work is done by agents.
- Market & Business
The largest US grid says it will cut power to big data centres during shortages
Electricity has stopped being an assumption and become a scheduled constraint - anyone planning capacity in that footprint now has an interruption window to design around.
- Market & Business✓ verified
More than 1,100 employees at four frontier labs asked the US government to build a way to pace AI development
Nothing about your model access, rate limits or roadmap changes this week; what is worth tracking is whether the international effort it asks for turns into a concrete proposed mechanism — reporting requirements, capability thresholds, a coordination body — because that is the point at which it stops being a signal and becomes a constraint you would plan around.
- AI in Education
New York puts 450 teenagers on an AI design challenge as summer-jobs programming
Paid summer employment is becoming an AI-literacy delivery channel, and it reaches the students an elective course never does.