Search

The Verge's reporting on a wave of rogue AI attack tests puts one company at the center of a story that now touches OpenAI, Meta, Anthropic, and Google. The important shift is not that agents can be prompted into risky behavior; it is that testing those behaviors has become a live operational discipline.

Developer ToolsUnscheduled

OpenAI's GPT-6 prompt caching update is really about production economics

OpenAI's prompt caching update for GPT-6 sounds like a developer feature, but the real story is cost control. Better cache hit rates, diagnostics, explicit breakpoints, and controls are the kind of details that determine whether AI workflows are affordable at scale.

ProductsSep 14, 2026

Perplexity's Astra case study shows AI search is becoming systems engineering

OpenAI's Perplexity case study is worth reading as a product-systems story, not a customer quote. Improving answer accuracy in AI search depends on retrieval, model behavior, evaluation, latency, and monitoring working together.

ResearchSep 14, 2026

Many-agent research is becoming a test bed for long-horizon reasoning

The arXiv work behind Stellar Colosseum points to a growing research pattern: instead of testing one model on one prompt, researchers are building many-agent environments where systems have to reason over longer horizons.

ProductsSep 11, 2026

LinkedIn's AI job-search work shows product AI needs training infrastructure, too

LinkedIn's AI job-search work is a reminder that useful AI products often depend on training systems most users never see. InfoQ's coverage of its multi-teacher approach shows how much engineering goes into matching people, jobs, and context at platform scale.

Policy and SafetySep 4, 2026

AI security teams are moving toward deeper red-team testing

AI safety debates can feel abstract until systems start acting in ways their builders did not expect. The next phase of red-team testing has to cover behavior over time, tool use, social engineering, and the ways agents behave when goals collide with boundaries.

AgentsSep 3, 2026

Agent memory poisoning turns persistence into a new security risk

AI agents are becoming more useful because they can remember. That same persistence creates a new security problem: if attackers can poison memory, they may influence future actions long after the original interaction is over.

ProductsSep 1, 2026

OpenAI is selling AI-native operations, not just better chat

OpenAI’s latest enterprise messaging is centered on workflows becoming operating capability. That is a useful shift because the real business value of AI is not a smarter prompt box; it is whether teams can redesign repeatable work around model-powered systems.

ResearchAug 31, 2026

Post-training is starting to look like maintenance work, not magic

A useful AI research signal this week is the move to describe LLM post-training as industrial maintenance. That framing is important because many model improvements depend less on mystery and more on cleaning, shaping, measuring, and repairing the data systems around the model.

ProductsAug 29, 2026

Local LLMs are becoming a practical privacy option for ordinary users

Running a chatbot on your own computer used to feel like a hobbyist project. It is becoming a practical option for people who want more privacy, lower recurring costs, or control over models that do not need to send every prompt to a remote service.

AgentsAug 28, 2026

OpenAI persistent agents would turn coding tools into always-on coworkers

A coding assistant that answers a prompt is easy to understand. A coding assistant that stays awake, notices unfinished work, and starts its own follow-up tasks is a much bigger bet. It turns software development from a request-response workflow into something closer to managing a tireless teammate.

ResearchAug 27, 2026

RedEvoAgent shows agent red-teaming is becoming its own automation race

As agents gain tool access, safety testing has to become more dynamic. Static prompt tests cannot fully capture systems that plan over time, use tools, and accumulate context across attempts.

ResearchAug 26, 2026

TraceML asks whether coding agents can plan through real ML work

Coding agents look impressive on isolated tasks, but machine-learning work is messier: data changes, experiments fail, metrics mislead, and progress often depends on choosing the next test rather than writing the next function. TraceML is useful because it studies that planning layer instead of treating every software task like a short coding puzzle.

Developer ToolsAug 26, 2026

Hugging Face's multi-vector encoder guide brings retrieval tuning closer to builders

Retrieval quality is still one of the quiet failure points in AI products. A model can be strong, but if the wrong documents reach the prompt, the answer looks confident and misses the point. Hugging Face's new multi-vector encoder material matters because it gives builders a more practical path to tune the retrieval layer itself.

AI in PracticeAug 26, 2026

Enterprise AI is moving toward data-local deployment patterns

Enterprise AI adoption is increasingly constrained by where the data lives. Companies want the productivity gains, but they do not want sensitive records, customer data, or regulated workflows flowing into systems they cannot govern.

InfrastructureAug 25, 2026

OpenAI's Jalapeno chip keeps inference efficiency in the spotlight

Jalapeno remains important because it points at the pressure underneath every AI product: serving prompts quickly, cheaply, and reliably. Model intelligence gets the headline, but inference economics decide how often users can actually use that intelligence.

Developer ToolsAug 24, 2026

SWE Refactor Bench tests whether coding agents can complete repository migrations

A benchmark focused on large-scale refactoring targets a practical question: can coding agents preserve behavior while changing many files?

Developer ToolsAug 23, 2026

Cloudflare uses AI to enforce engineering standards

InfoQ reports on Cloudflare using AI to enforce engineering standards, a concrete example of AI moving into software delivery governance.