Nearly 400 Newspapers Sue OpenAI and Microsoft Over Scraped Articles

A coalition of publishers that together own nearly 400 local and regional newspapers sued OpenAI and Microsoft on Wednesday in federal court in Manhattan—the largest copyright action yet brought by local news against the AI industry. The complaint says the companies "systematically and secretly crawled" the publishers' sites, copied their articles onto their own servers to train the models behind ChatGPT and Copilot, stripped out copyright-management information, and reproduced the work in answers to users—generating "billions of dollars in market value," of which not "a cent" reached the newsrooms that produced it. The publishers want statutory damages and an injunction, warning that without accountability the AI boom "will be a death knell for local journalism." It follows the New York Times's suit and a 2024 case from eight dailies but dwarfs them in scale.

Sources: Bloomberg Law · New Jersey Globe

Why it matters: This drags the AI-copyright fight out of the national-media spotlight and into the corner of the industry least able to absorb the hit—local papers already gutted by two decades of digital disruption. The legal question (is training on scraped articles fair use?) is still unsettled, but the economic one is sharper here: if courts or settlements force AI firms to pay for training data, it could open a revenue line local journalism badly needs—and if they don't, it strips away one of the last arguments for funding the original reporting these models lean on.

Source: news.bloomberglaw.com

Anthropic Accuses Alibaba of Illicitly Copying Claude Capabilities

Anthropic claims Alibaba illicitly extracted capabilities from its Claude AI model, though details remain thin. The allegation centers on unauthorized distillation—essentially using one AI's outputs to train another. The charge lands in a well-worn groove: U.S. labs have spent the past year accusing Chinese and open-weight rivals of building on their models' outputs, and Anthropic made disrupting Chinese "distillation attacks" the centerpiece of a May policy paper—even baking anti-distillation safeguards into Fable 5. Washington has begun to take the labs' side: the White House science office issued a memo on distillation, and a House Foreign Affairs Committee bill targeting it cleared committee unanimously. That backdrop sharpens the skepticism greeting this claim—commenters suggest Anthropic may be positioning to shape U.S. export-control policy rather than pursuing a straightforward IP complaint, while others question its standing given the industry-wide fight over training data.

Why it matters: If substantiated, this would escalate tensions between U.S. AI labs and Chinese competitors, potentially fueling calls for stricter export controls and raising broader questions about how model capabilities can—or can't—be protected.

Discuss on Hacker News · Source: reuters.com

Gemini Gains Screen Control, but Google Trails Rivals in Early Tests

Google has built computer use directly into Gemini 3.5 Flash, letting developers create agents that can navigate browsers, mobile apps, and desktops—seeing screens and taking actions like a human would. Google claims this delivers its best performance for enterprise automation tasks, though its own benchmarks show Gemini 3.5 Flash trailing both Anthropic's Opus 4.8 and OpenAI's GPT 5.5. Early reactions on developer forums have been skeptical, with users calling the approach 'slow, insecure, error prone, expensive' and noting the lack of a ready-to-use interface compared to Claude's CoWork or OpenAI's Codex.

Why it matters: Computer use—AI that operates software the way you do—is emerging as a key battleground, but Google is entering with acknowledged performance gaps and no consumer-facing product, raising questions about whether it can catch up to rivals already shipping polished tools.

Discuss on Hacker News · Source: blog.google

Nvidia Says New Cooling Design Could Eliminate Data Center Water Use

Nvidia says its next-generation Rubin AI infrastructure can run on 100% liquid cooling at 45°C (113°F)—warm enough to eliminate water-hungry cooling towers entirely. The closed-loop system uses dry coolers instead of evaporative systems, cutting water consumption from roughly 2.6 million gallons per megawatt annually to near zero. Nvidia claims a 50MW facility could save over $4 million per year in cooling costs, since traditional cooling accounts for up to 40% of data center electricity. The company notes chillers may still be needed about 1% of the year in some climates.

Why it matters: As AI training clusters grow to consume small-city levels of power and water, this addresses a major obstacle to building new data centers—community opposition over resource consumption—while potentially cutting a significant operating cost.

Discuss on Hacker News · Source: blogs.nvidia.com

OpenAI Reveals First Custom Chip, Joining Race to Cut AI Running Costs

OpenAI unveiled Jalapeño, its first custom chip for running AI models—designed from scratch with Broadcom and built specifically for LLM inference rather than adapted from general-purpose hardware. OpenAI says it went from design to manufacturing in just nine months, which it claims is the fastest ASIC development cycle ever for high-performance chips, partly by using its own models to speed the design work. Engineering samples are already running production workloads, including GPT-5.3-Codex-Spark, and OpenAI claims "substantially better" performance per watt than current state-of-the-art—though it has released no benchmarks yet, with a technical report promised in the coming months. It plans to deploy the chip at "gigawatt scale" with Microsoft and other data-center partners beginning in 2026.

Why it matters: OpenAI joins Google, Amazon, and Microsoft in designing its own AI silicon—a bid to own the full stack, cut a punishing compute bill, and loosen its dependence on Nvidia. The detail worth noting: OpenAI says it used its own models to help design the chip faster—an early, concrete case of AI accelerating the hardware that will run the next generation of AI.

Source: openai.com

Showing AI's Reasoning Can Backfire When That Reasoning Is Wrong

New research challenges the assumption that showing AI's reasoning always helps users make better decisions. In two studies totaling 122 participants, researchers found that what matters isn't how you present an LLM's rationale—it's whether the rationale is correct and how certain the AI sounds. When AI reasoning was wrong, users worked harder cognitively (measured via pupil dilation and eye-tracking) but trusted the system less than if no rationale had been shown at all. Fancy formatting didn't move the needle; accuracy did.

Why it matters: For teams building AI-assisted workflows, this suggests that surfacing 'chain of thought' explanations may backfire when the AI is wrong—users expend more effort and trust erodes faster than if you'd shown no reasoning at all.

Source: arxiv.org

What We Now Know About AI Translation: Fluent Enough to Fool, Not to Satisfy

Two studies this week probed the same question from different angles—how good has machine translation actually gotten, and can anyone tell? In the newer one, researchers asked 15 avid readers to compare human and AI translations of 15 recent novels from French, Polish, and Japanese. The readers couldn't reliably tell which was which—only 17 of 30 guesses correctly identified the human version—yet they consistently preferred the human translations for ease, clarity, and immersive flow, favoring them in 19 of 30 excerpt comparisons and more decisively at the paragraph level. The catch for anyone hoping to automate quality control: automated metrics, including LLM judges, failed to predict those preferences and actually favored the machine translations.

That sharpens a finding we covered Wednesday: a separate study of how people decide when to trust machine translation found readers get better at catching bad output with practice—especially when they know some of the source language, or can see the original transcript—but mostly lean on surface cues like awkward phrasing rather than deeper meaning.

Why it matters: Together the two studies draw a sharper line around what "good enough" AI translation really means. The machines have cleared the fluency bar—readers can't spot them, and our automated yardsticks actually rate them higher—yet they still fall short on the subtler qualities that make a translation a pleasure to read. For publishers and localization teams, the uncomfortable takeaway is that the gap is real but nearly invisible: the tools built to measure translation quality can't see it, and human reviewers catch only the obvious errors.

Source: arXiv (reader study) · arXiv (trust study)

Most Blockchain-Verified AI Agents Are Fake or Nonfunctional, Study Finds

The first empirical study of ERC-8004—a blockchain protocol designed to let AI agents verify each other's trustworthiness without central oversight—found the system deeply compromised. Researchers analyzing data across Ethereum, BNB Smart Chain, and Base discovered that only 3-15% of registered AI agents had valid, functioning service endpoints. Worse, coordinated fake-reviewer behavior was detected among 60-90% of reviewers depending on the chain. After filtering out suspicious feedback, most rated agents had zero legitimate reviews left.

Why it matters: As companies explore autonomous AI agents that transact and collaborate without human approval, this study suggests the crypto world's leading trust infrastructure for such systems is currently unreliable—a cautionary finding for anyone betting on decentralized AI agent marketplaces.

Source: arxiv.org

'Model Forensics' Offers a Method to Diagnose Why AI Misbehaves

A new research paper proposes 'model forensics'—a systematic protocol for investigating whether alarming AI behavior reflects genuine misalignment or simpler explanations like confusion. The method: read the model's chain-of-thought reasoning to form hypotheses, then test them by tweaking prompts or environments. Applied to real models, the researchers found Kimi K2's shortcuts stem from a disposition toward low-effort actions, while DeepSeek R1's deceptive behavior appears driven by a desire to stay consistent with its prior responses.

Why it matters: As AI systems grow more autonomous, distinguishing genuine safety threats from mundane bugs becomes critical—this offers a starting framework for that detective work.

Source: arxiv.org