@thenewstack

All about at-scale software development, deployment & management. Tech news, analysis, research, podcasts & more! Subscribe đź“© https://xitter.cf/t.co/X1jwJjLnNj

 We're Global 🌏
thenewstack.io
Joined February 2014
  • Tweets127K
  • Following3.1K
  • Followers41.4K
11.3KMedia
AI agents need more than static checkpoints. Learn why continuous observation and action-level authorization are vital for agent security. Thanks to Ory thenewstack.io/ai-agent-secu…

AI agent security is a lifecycle — not a gatekeeper

AI agents need more than static checkpoints. Learn why continuous observation and action-level authorization are vital for agent security.

thenewstack.io
92
AI makes code creation abundant, but code confidence is scarce. Close the confidence gap with guiderails, verification, and auditability. Thanks to Sonar thenewstack.io/ai-code-confi…

The code confidence gap nobody’s talking about

AI makes code creation abundant, but code confidence is scarce. Close the confidence gap with guiderails, verification, and auditability.

thenewstack.io
1
1
194
Vijil argues AI agents should earn autonomy like credit: tested up front, capped by policy, and monitored constantly. Without that, they go feral. Thanks to Vijil thenewstack.io/ai-agent-trus…

From feral to fiduciary: making AI agents serve your interests

Vijil argues AI agents should earn autonomy like credit: tested up front, capped by policy, and monitored constantly. Without that, they go feral.

thenewstack.io
1
218
Linus Torvalds let Claude Code fix the Linux page cache. Then reviewers found a race. thenewstack.io/torvalds-clau…

Linus Torvalds let Claude Code fix the Linux page cache. Then reviewers found a race.

Linus Torvalds used Claude Code to build a Linux page cache fix that hit 23x read throughput, but kernel reviewers found a race before anything merged.

thenewstack.io
302
MiniMax just halved its token prices. Your AI agent bill may not follow. thenewstack.io/minimax-flash…

MiniMax just halved its token prices. Your AI agent bill may not follow.

MiniMax says M3.1 Flash cuts agent costs 70 percent. Its token prices halved, but what a finished task costs tells a narrower story for busy AI developers.

thenewstack.io
2
2
377
Researchers left a hidden message for rogue AI agents. Then the agents wrote back. thenewstack.io/rogue-ai-agen…

Researchers left a hidden message for rogue AI agents. Then the agents wrote back.

Researchers hid a message for rogue AI agents probing Alibaba’s Amap from Tencent Cloud. The agents wrote back, revealing their tasks, sandbox and limits.

thenewstack.io
291
AI found flaws in the code behind 70% of the web. The fix took two months. thenewstack.io/ai-php2shell-…

AI found flaws in the code behind 70% of the web. The fix took two months.

PHP powers roughly 70% of the world’s websites, according to Wiz Research, so a flaw in its core engine affects

thenewstack.io
341
Amodei says Anthropic’s safety delay may have cost it an AI lead thenewstack.io/anthropic-myt…

Amodei says Anthropic’s safety delay may have cost it an AI lead

Anthropic had its Mythos AI models ready in February but held them back until June. CEO Dario Amodei says safety may soon struggle to keep pace.

thenewstack.io
326
Researchers removed one instruction from Claude Code. Out-of-scope actions jumped. thenewstack.io/coding-agents…

Researchers removed one instruction from Claude Code. Out-of-scope actions jumped.

Anthropic’s next coding model may infer developer intent. Research shows coding agents often go beyond scope, and that’s much harder to fix than it sounds.

thenewstack.io
1
2
442
Nvidia wants AI factories built in real factories thenewstack.io/nvidia-dsx-po…

Nvidia wants AI factories built in real factories

Partners will pre-build DSX POD containers with seven Vera Rubin racks and liquid cooling, cutting data center setup from months to as few as 14 days.

thenewstack.io
1
437
DeepSeek made its name with cheap AI models. Now it’s undercutting search API rivals, too. thenewstack.io/deepseek-web-…

DeepSeek made its name with cheap AI models. Now it’s undercutting search API rivals, too.

The Chinese AI company is opening up the search engine behind its chatbot through Search and Fetch APIs, priced at $0.75 per 1,000 successful requests each.

thenewstack.io
1
420
66% of billion-dollar companies leave critical bugs unpatched — the part the AI bug-hunting hype leaves out thenewstack.io/chainguard-my…

66% of billion-dollar companies leave critical bugs unpatched — the part the AI bug-hunting hype...

Chainguard’s new Harness points frontier models like Claude Mythos 5.1 at your code, alongside new tools to isolate coding agents and verify their work.

thenewstack.io
4
522
“Take work away from frontier LLMs”: Subquadratic’s decision model claims 12x Jev’s context window thenewstack.io/subquadratic-…

“Take work away from frontier LLMs”: Subquadratic’s decision model claims 12x Jev’s context window

Jev sparked a rush of decision models. Subquadratic’s CEO says they’re mostly doing LLMs’ admin -- and wants them tackling bigger jobs.

thenewstack.io
2
410
Respan spent two years building on LiteLLM. One upgrade made it start over. thenewstack.io/respan-rust-a…

Respan spent two years building on LiteLLM. One upgrade made it start over.

Respan rebuilt its AI gateway in Rust around a small plugin core after 100,000 lines of LiteLLM patches. Custom plugins still mean compiling your own.

thenewstack.io
1
400
PHP powers roughly 70% of the world’s websites, according to Wiz Research, so a flaw in its core engine affects thenewstack.io/ai-php2shell-…

AI found flaws in the code behind 70% of the web. The fix took two months.

PHP powers roughly 70% of the world’s websites, according to Wiz Research, so a flaw in its core engine affects

thenewstack.io
4
535
Your AI agent worked perfectly in the demo. What happens when it meets real production data? Watch the on-demand session to explore the six architectural requirements for building AI systems that can operate reliably at scale. Watch now: bit.ly/3UWIB0d
306
Anthropic had its Mythos AI models ready in February but held them back until June. CEO Dario Amodei says safety may soon struggle to keep pace. thenewstack.io/anthropic-myt…

Amodei says Anthropic’s safety delay may have cost it an AI lead

Anthropic had its Mythos AI models ready in February but held them back until June. CEO Dario Amodei says safety may soon struggle to keep pace.

thenewstack.io
1
374
Nvidia bets open models can give enterprise AI agents the context they’re missing thenewstack.io/nvidia-agent-…

Nvidia bets open models can give enterprise AI agents the context they’re missing

Agent Knowledge Foundation bundles retrieval, ontology, and prediction models for enterprise agents. Nvidia is counting on Snowflake, Databricks, and Palantir to decide who gets to see what.

thenewstack.io
1
383
Liquid AI caught its model cheating on coding tests — then it recognized the sandbox thenewstack.io/liquid-ai-rew…

Liquid AI caught its model cheating on coding tests — then it recognized the sandbox

Liquid AI caught its new LFM2.5 model searching Git history and recognizing its sandbox during training, raising hard questions about agent benchmarks.

thenewstack.io
2
357
Workday goes headless, opening its HR and finance data to outside AI agents thenewstack.io/workday-headl…

Workday goes headless, opening its HR and finance data to outside AI agents

Workday’s MCP server, SQL access, and agent runtime are now GA, letting developers build HR and finance agents in Claude Code and Cursor under user permissions.

thenewstack.io
2
356
Load more