China’s producer inflation swung higher last month, as escalating hostilities in the Middle East sent oil prices surging past $100 a barrel and goods related to AI became more expensive.

U.S. authorities warn that sanctioned hackers are increasingly targeting Americans to act as facilitators between breached U.S. companies and Pyongyang.

Discussion | Link

Wharton Human-AI Research and GBK Collective’s new report, The Adaptive Advantage: Who’s Winning with AI and How, shows how far enterprise AI has spread—and how much harder it is to turn adoption into substantial business returns. Eighty percent of managers and more senior respondents report positive returns on their department’s AI investments, but only 28% describe those returns as strongly positive. The fourth annual study points to workflow redesign, clearer ownership, and growth-oriented…

Bloomberg: Filing: Abdulrahman Alfozan, the OpenAI staffer charged with bugging a conference room, used a Seeed Xiao ESP32-S3 Sense dev board; a website sells it for $14  —  An employee at OpenAI was arrested and charged with using a hidden recording device in one of the company's San Francisco offices …

Rashi Shrivastava / Forbes: Albert Invent, which launched an AI model trained on chemical patents and data about 15M molecules, raised a $55M Series B, a source says at a $340M valuation  —  Scientists at manufacturing conglomerates like Kenvue and Henkel use Albert Invent's AI tools to design experiments.

New analysis shows how agentic AI adoption is growing across EU enterprises and small businesses—and where deeper use could unlock more value.

Lisbon , Portugal - 12 November 2025; Jesse Zhang, Co-founder & CEO, Decagon AI on Centre Stage during day two of Web Summit 2025 at the MEO Arena in Lisbon, Portugal. (Photo By Shauna Clinton/Sportsfile for Web Summit via Getty Images) Shauna Clinton/Sportsfile for Web Summit via Getty Images AI startup Decagon builds AI agents to automate customer support for enterprise customers. The company has hired an investment bank to evaluate M&A interest. There have been early talks with potential buyers. Update: This story has been updated to reflect that ServiceNow says it is no longer in talks with Decagon. Buzzy AI agent startup Decagon could be the latest company swept up in a growing wave of AI dealmaking. The company, which was last valued at $4.5 billion earlier this year, has hired a bank to evaluate acquisition interest, according to people familiar with the matter. Decagon held early discussions with cloud software giant ServiceNow, but a spokesperson for ServiceNow says they're not ongoing. People familiar with the matter told Business Insider there have also been early talks with other companies; if a deal does not materialize, Decagon could instead raise a new funding round. "Our approach to M&A has not changed: we remain an organic growth company first. M&A is a tool to accelerate our roadmap or open meaningful new innovation," said Amit Zavery, president, chief product officer, and chief operating officer of ServiceNow. The interest in Decagon represents a new shift in the AI boom: after years of huge venture rounds and soaring valuations, tech companies are increasingly looking to buy their way into promising AI markets. Hugging Face was recently acquired by Nvidia for 12.9 billion, and OpenRouter was snapped up by Stripe for around $8 billion. Growing AI-related businesses are often getting offers at a significant premium to their last valuation, so if Decagon were to accept an offer, it could be considerably higher than its previous $4.5 billion value. Founded in July 2023 by Jesse Zhang and Ashwin Sreenivas, Decagon uses generative AI to automate customer support for enterprise customers, including Delta, Snapchat, and Ticketmaster. The company has raised over $400 million from a roster of well-known Silicon Valley investors, including Andreessen Horowitz, Index Ventures, Bain Capital Ventures, and Accel. It competes with industry veteran Bret Taylor's Sierra, which was valued at $15.8 billion earlier this year. ServiceNow's previous interest in Decagon came as it looked to reassure investors that it can adapt to AI’s threat to traditional software businesses. Its shares tumbled earlier this year during the so-called SaaS-pocalypse, when investors feared AI could erode software companies’ profit margins. It was also punished for making a $7 billion acquisition of the security business Armis, a deal announced late last year. The stock has since rebounded by 55% in the past six months as ServiceNow continues to invest heavily in AI and reshape parts of its business. Read the original article on Business Insider

AI-powered knowledge management systems face a fundamental challenge when businesses need accurate answers rather than plausible ones. Imago Platform Ltd. encountered that problem while developing an application to help customers navigate complicated product warranty policies. Imago’s first client, a lighting and electrical business, needed help resolving warranty claims across manufacturers with widely varying policies. The […] The post Imago Platform uses graphs to improve AI knowledge management appeared first on SiliconANGLE.

The Information: Sources: OpenAI is in early talks to buy Trail of Bits, which finds and fixes vulnerabilities in open-source software; founded in 2012, it has 40+ employees  —  OpenAI, which is trying to develop a substantial cybersecurity business, is in early talks to buy Trail of Bits …

Chatbot prompts reveal the engineer’s alleged efforts to set up a covert listening device days before one was found in a meeting room used by high-level executives.

Max Hodak, the former president and co-founder of Neuralink, detailed onstage at TechCrunch Disrupt how his company's work with PRIMA is part of its path toward a novel brain-computer interface (BCI).

Physical Intelligence announced the Pi API beta on October 13, 2026, making its latest robotic foundation models available to outside developers for hosted fine-tuning and inference. Access to the beta runs through a waitlist, with applicants asked to describe what they want to build. Under the workflow laid out in the announcement, a developer uploads robot data, launches fine-tuning on compute and infrastructure that Physical Intelligence manages, and accesses the fine-tuned model through a…

Cerebras CEO Andrew Feldman predicts a future where personal AI assistants become as commonplace as household appliances, running partly on local devices and partly in the cloud.

Anthropic on October 12, 2026, detailed expanded and strengthened defenses against distillation attacks targeting Claude, saying early evidence suggests the new methods have reduced the amount of data harvested by distillers by around 90 to 99 percent. In the post, Strengthening our defenses against distillation attacks, the company defined distillation attacks as covert, industrial-scale campaigns to extract one model's capabilities and replicate them in another without authorization, often…

Reuters: Tencent raised ~$5.7B via US dollar and offshore yuan debt offerings, drawing $23B+ in orders; source: it had initially been expected to raise at least $4.5B  —  Tencent Holdings (0700.HK) has raised about $5.7 billion through US dollar and offshore yuan debt offerings on Tuesday …

<p><strong><a href="https://www.anthropic.com/engineering/redesigning-source-control">Redesigning source control at Anthropic</a></strong></p> Interesting piece from Anthropic describing why and how they built their own custom alternative to Git, based on the dramatic increase in rate of changes to their code after the past six months.</p> <blockquote> <p>In May 2026, Anthropic averaged around 2,500 code changes merged per day; as of September, we merged over 30,000 per day.</p> </blockquote> <p>They included a chart showing those changes. I <a href="https://claude.ai/share/aefcdd4a-9364-40c1-927a-29bf445894fc">had Claude modify their chart</a> (<a href="https://gist.github.com/simonw/4a16dff066fd9763faf8642f797905be">code here</a>) to add markers for the public releases of each of their models in 2026:</p> <p><img alt="Stacked bar chart titled “Changes merged per day”, for the Anthropic monorepo, January–September 2026. Each day’s bar is split into pull requests (blue) and antique change requests (orange), with a black line for the 7-day average. Weekdays start at around 1,000 merges in January and climb slowly to about 3,000 by late May. Antique change requests appear in late May, overtake pull requests in mid-June and have almost fully replaced them by mid-July, when weekdays reach 6,000–8,000. A labelled “Quiet week” at the end of July dips below 1,000 a day. After that, growth turns steep: over 10,000 a day by mid-August, about 17,000 in early September, and a peak of 30,401 on Sep 9. Dotted lines mark Claude model releases: Opus 4.6 (Feb 5), Sonnet 4.6 (Feb 17), Mythos Preview (Apr 7), Opus 4.7 (Apr 16), Opus 4.8 (May 28), Fable 5 and Mythos 5 (Jun 9), Sonnet 5 (Jun 30), Opus 5 (Jul 24, just before the quiet week), and Fable 5.1 and Mythos 5.1 (Sep 1)." src="https://static.simonwillison.net/static/2026/anthropic-merges-per-day.webp" /></p> <p>This shows public release dates, not the date when models became available for Anthropic's internal use. My guess is that the introduction of Fable/Mythos class models is a key factor in the increased coding velocity.</p> <p>"Antique" is their new code review system. The source control system is called Scion, and sounds to me like a reinvention of centralized source control along the lines of Subversion.</p> <p>(<strong>Update</strong>: Anthropic credit Google's <a href="https://en.wikipedia.org/wiki/Piper_(source_control_system)">Piper</a> and Meta's <a href="https://sapling-scm.com/">Sapling</a> as "prior art" - it looks like the core concept of "derived data" <a href="https://sunshowers.io/posts/anthropic-mononoke/">was heavily inspired by Mononoke</a>, the server component of Sapling.)</p> <p>They had some pretty extreme requirements:</p> <blockquote> <p>The history of our main branch alone is almost 100 GB, and a <em>compressed</em> working copy is over 25 GB. We perform on the order of 25 million code checkouts per day for CI, before counting developer environments, agent sandboxes, and research jobs. Fetched naively, that would be over 600 PB of git traffic per day.</p> </blockquote> <p>The new system was built by a team of seven people, and the migration from Git was completed in five months.</p> <p>When asked on Hacker News about the lack of offline support, Scion engineer Nick G <a href="https://news.ycombinator.com/item?id=50073633#50079495">replied</a>:</p> <blockquote> <p>Since Claude is by far the heaviest user of Scion and is necessarily online, this just wasn't important to do.</p> </blockquote> <p>As an aside, I've been trying adversarial review in my own projects recently, where you have a model try and find flaws (I often play Claude and GPT off against each other). This note from the article reflected my own experience with that:</p> <blockquote> <p>Anthropic also makes extensive use of adversarial review, and, while this is incredibly effective in many ways, it can also be myopic, so skilled engineers needed to steer Claude away from over-engineering parts of the system in response to hypothetical concerns.</p> </blockquote> <p><small></small>Via <a href="https://news.ycombinator.com/item?id=50073633">Hacker News</a></small></p> <p>Tags: <a href="https://simonwillison.net/tags/git">git</a>, <a href="https://simonwillison.net/tags/migrations">migrations</a>, <a href="https://simonwillison.net/tags/version-control">version-control</a>, <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/generative-ai">generative-ai</a>, <a href="https://simonwillison.net/tags/llms">llms</a>, <a href="https://simonwillison.net/tags/anthropic">anthropic</a>, <a href="https://simonwillison.net/tags/agentic-engineering">agentic-engineering</a></p>

Industrial Next helps manufacturers automate complex, delicate tasks that have traditionally required human labor.

"I can always find a couple of biotech stocks to smoke," Shkreli said in a livestream.

Getty Images Next month, ChatGPT will mark its fourth birthday. In that short time, the popular chatbot and other generative artificial intelligence (GenAI) tools have transformed society as we know it. Tech leaders have promised GenAI will unlock new levels of economic productivity and human prosperity. Critics warn it could instead deepen inequality. Whatever its long-term effects, the pressure to adopt it is growing. New Zealand’s government expects agencies to use GenAI to create a leaner, more efficient public service. Companies are pushing “AI-first” initiatives and spending heavily on GenAI services to make sure they aren’t left behind. Most New Zealanders have now had some interaction with this technology, with more than half of us using it at least weekly. Some are trying to limit their use because of concerns about environmental costs or the impact on creative industries. But avoiding AI altogether is becoming increasingly difficult as it spreads through search engines, advertising and online content. And while the current AI boom may eventually slow, GenAI is unlikely to disappear any time soon. So how should New Zealand, a small nation far from the major economies hosting AI giants, learn to live with it? The next government has more choice than we might think. Australia and other “middle powers” have already introduced controls to manage AI’s risks. With the election approaching, and parties pitching a range of policies, New Zealand is overdue a serious debate about what safeguards it needs. NZ’s ‘light-touch’ approach The government’s 2025 AI strategy deliberately takes a “light-touch approach”, encouraging AI adoption while relying largely on existing laws to manage its risks. Rather than introducing specific AI regulations, it favours voluntary guidelines and frameworks that organisations are encouraged, but not required, to follow. This has raised concerns about whether legal protections can keep pace. There is also public appetite for stronger action, with one survey finding eight in ten Kiwis believe AI regulation is needed. Another study found just 5% of respondents were strongly confident the government would use AI responsibly. There are particular concerns about government agencies’ use of automated decision-making (ADM) systems, most of which are built on machine learning techniques that predate GenAI. Done well, these systems can make life easier for people. ACC claims can be automatically approved before patients leave their GP’s office, while tax refunds can be paid without taxpayers having to take any action. But done badly, it can make headlines. In 2018, Immigration New Zealand used an ADM model to identify overstayers for possible deportation, prompting accusations of racial profiling. In Australia, the Robodebt scheme issued unlawful welfare debts, forcing the government to repay A$721 million. Read more: Election Policy Tracker: your guide to what the parties are promising during the 2026 election campaign Yet New Zealand’s laws provide limited oversight of the technology. There are some protections under the Official Information and Privacy Acts, but they’re arguably unfit for purpose. The latest GenAI systems use even more complex models, making it harder to identify problems such as bias and discrimination, which risks disproportionately affecting Māori and people with disabilities. Meanwhile, parliament recently passed a law allowing the Ministry of Social Development to automate welfare benefit decisions. The Law Commission is presently developing a framework to govern ADM across government. Such efforts are overdue: Canada’s government began requiring its agencies to assess the impacts of such systems nearly seven years ago. Who should benefit from AI? Regulation is only part of the challenge. New Zealand also needs to consider who controls – and who benefits from – the AI systems it increasingly relies on. Many of these systems, including those used for ADM, are developed and hosted overseas. This has implications for the protection of personal information, intellectual property and New Zealand’s ability to retain control over important public services. Tech companies’ use of copyrighted material to train AI models has already prompted disputes over whether creators should be compensated. There are other big questions about where the money goes. Much of New Zealand’s spending on AI flows to international technology companies, some of which declare relatively little income here. While many AI services are not yet profitable, their developers are betting on substantial future returns. New Zealand needs to consider how those profits should be taxed before they materialise. The government might argue these giants are too difficult to tax. Critics would counter this simply reflects a lack of political leadership. Arguments also extend to AI infrastructure. Data centres, such as one proposed for Southland, have received plenty of public attention. Their heavy energy demands and environmental impacts indeed deserve scrutiny, but they’re only one part of the debate about how New Zealand can benefit economically and socially from AI. Other countries offer possible models for addressing these wider issues. Australia has established an AI Safety Institute, the European Union has introduced risk-based AI regulations and Canada requires government agencies to assess the impacts of automated decisions. But New Zealand has its own obligations to consider and any approach must also reflect te Tiriti o Waitangi. That means considering Māori rights and interests and ensuring Māori involvement in designing and governing AI systems, particularly where Māori data is concerned. Māori principles of data sovereignty emphasise community control over information and the benefits derived from it. As an example, the non-profit Te Hiku Media has already put these principles into practice, developing language technologies that keep control of te reo Māori data within the community. These approaches show that governments and communities can have a greater say in how AI is developed and used. Whoever is elected to power next month, it’s clear New Zealand need not leave these decisions entirely to overseas technology companies. From the cost of living to how best to tackle climate change, the coming election involves some big choices. This article is part of The Conversation’s Election 2026 series: expert analysis of the major policy debates and the big challenges New Zealand’s next government will face. Andrew Lensen receives research funding from the Ministry of Business, Innovation and Employment and the Royal Society of New Zealand. He is a co-director of LensenMcGavin AI, a consultancy specialising in the responsible uptake of AI in Aotearoa.