AI Arena
Data Science
Can an AI catch data leakage before you trust the score?
A new experiment plants four common leaks in real pipelines and asks Claude models to spot them.
Read the story →Archive
All stories.
The full archive of AI Arena briefings.
Policy
Kanter's new startup wants to put antitrust AI in plaintiffs' hands
The former DOJ antitrust chief has a new firm, FluxCap, built to help smaller legal teams take on corporate giants.
Read the story →
Products
Automattic launches Spacefast for vibe-coded apps
The WordPress owner is betting that AI-built sites need a place to run, not just a prompt box.
Read the story →
Products
Apple is pushing Siri to do more than talk
A reported roadmap puts deep app control at the center of Siri's next chapter, as Apple tries to catch the agent leaders.
Read the story →
Lawsuits
Elvis and Eminem rights holders take Suno to court
Two lawsuits say the AI music company reproduces voices and lyrics without permission — and one leans on Tennessee's ELVIS Act.
Read the story →
Engineering
Giving an AI agent memory that survives the session
A tutorial pairs the OpenAI Agents SDK with Mem0 to store preferences and retrieve them later.
Read the story →
Research
A 32MB voice model that runs in the browser
Underdog's Woof demo turns a retro game into a test of what fits on a local device.
Read the story →
Data Science
Seven Python libraries for the time series stack
A tour of the tools that cover statistical forecasting, machine learning pipelines, deep learning, and anomaly detection.
Read the story →
Robotics
What a $30 robot cleaner is really selling
Tau Robotics sends remote-piloted humanoids into San Francisco homes, and every wiped counter helps train its models.
Read the story →
Research
A benchmark for the messy work of wet labs
OpenAI and Ligo test whether AI agents can read gels, plates, and microscope images without overconfidence.
Read the story →
Business
Airbnb will launch its travel agent in early 2027
Brian Chesky says trip planning needs more than a chat box, so Airbnb is building its own interface.
Read the story →
Education
Australian universities push AI into the curriculum
New partnerships with OpenAI mix broad access with rules for when students should not use AI.
Read the story →
Products
Gemini Omni 1.2 Flash adds 15-second video and a 30% price cut
Google's new video model focuses on longer clips, consistent references, and edits that do not require a full re-render.
Read the story →
Security
OpenAI's GPT-6 Cyber finds real exploits, stays in scope
A security model that scores higher on attack benchmarks also learns to respect the boundaries researchers set.
Read the story →
Infrastructure
Texas voters just put the AI data center boom on notice
A Democratic sweep has the state's biggest buildouts facing permits, moratoriums, and a lot of uncertainty.
Read the story →
Research
Can better rules, not better code, keep AI agents in line?
Researchers are borrowing an idea from economics: reward the behavior you want instead of trying to read the model's mind.
Read the story →
Education
Google's new AI Quest teaches students to map a brain
The third lesson in the AI Quests series uses neuroscience to show where human judgment still matters.
Read the story →
Models
Gemini 4 Argon targets deep work, not chat
Google's new frontier model is priced for enterprises and tuned for coding, science, and long multi-step tasks.
Read the story →
Business
Jev, Laya, and Kev test a narrower kind of AI
Instead of writing the email, these models decide whether a quote request should be sent, held, or escalated.
Read the story →
Engineering
A local invoice agent that reads, decides, and flags
Surya 2 and Chandra 2 handle OCR on the laptop, while a small model decides whether to pay, file, or escalate.
Read the story →
Business
Why small local models beat the frontier for some jobs
A practical case for one 0.5–3B model doing one job on hardware you already own.
Read the story →