The week in AI

AI startup Recursive emerges from stealth with $650 million to build self-improving AI

AI startup Recursive has officially emerged from stealth, calling recursive self-improvement the "fastest path to superintelligence." The article AI startup Recursive emerges from stealth with $650 million to build self-improving AI appeared first on The Decoder.

Google is hiring hundreds of engineers to help customers adopt its AI

Google is hiring now also hundreds of engineers to help customers adopt its AI. A sign that implementation remains difficult. The article Google is hiring hundreds of engineers to help customers adopt its AI appeared first on The Decoder.

From Prompt to Pointer Engineering: Deepmind tries to reinvent the mouse cursor for the AI era

Pointer Engineering: Deepmind wants to turn the mouse cursor into the key variable in context engineering. The article From Prompt to Pointer Engineering: Deepmind tries to reinvent the mouse cursor for the AI era appeared first on The Decoder.

Android gets AI agents that book trips, fill forms, and clean up your texts

With Gemini Intelligence, Google is introducing new AI features for Android that automate multi-step tasks, summarize web content, fill out forms, and turn spoken thoughts into polished text messages. The article Android gets AI agents that book trips, fill forms, and clean up your texts appeared first on The Decoder.

Anthropic expands legal AI offerings with new Claude Cowork plugins

Anthropic launches twelve new Claude plugins for legal work covering contract law, employment law, and litigation, while connecting the chatbot to services like Thomson Reuters' CoCounsel Legal and Harvey. According to Anthropic's chief legal officer, lawyers use Claude more than almost any other profession. The article Anthropic expands legal AI offerings with new Claude Cowork plugins appeared first on The Decoder.

Google says it stopped a mass cyberattack after AI was used to discover a zero-day exploit

Google's Threat Intelligence Group has identified the first known case of an attacker using AI to discover and weaponize a zero-day vulnerability. Google says it stopped the planned mass attack. State-backed actors from China, North Korea, and Russia are also using AI to find vulnerabilities and disguise malware code. The article Google says it stopped a mass cyberattack after AI was used to discover a zero-day exploit appeared first on The Decoder.

Alphabet's Isomorphic Labs raises $2.1 billion to scale AI drug discovery toward clinical trials

Isomorphic Labs, the AI drug research company led by DeepMind co-founder Demis Hassabis, has closed a $2.1 billion Series B round led by Thrive Capital. The funding will go toward expanding its in-house platform IsoDDE and advancing drug candidates toward clinical trials. The article Alphabet's Isomorphic Labs raises $2.1 billion to scale AI drug discovery toward clinical trials appeared first on The Decoder.

Microsoft ousts its Israel chief following reports that Azure quietly powered military AI targeting in Gaza

Microsoft Israel's top executive is out after an internal investigation into the unit's work with Israel's defense ministry. Reporting over the past years points to what's likely at the center of it all: cloud infrastructure, mass surveillance, and AI-powered target selection in Gaza. The article Microsoft ousts its Israel chief following reports that Azure quietly powered military AI targeting in Gaza appeared first on The Decoder.

How finance teams use Codex

See how finance teams can use Codex to build MBRs, reporting packs, variance bridges, model checks, and planning scenarios from real work inputs.

"Tokenmaxxing" spreads at Amazon as employees game internal AI leaderboards

Amazon employees are automating unnecessary tasks just to climb internal AI leaderboards. The article "Tokenmaxxing" spreads at Amazon as employees game internal AI leaderboards appeared first on The Decoder.

Using Polars Instead of Pandas: Performance Deep Dive

In this article, we explore three real data problems using real questions where Polars outpaces Pandas on every metric.

Thinking Machines Lab ships its first model and argues interactivity is what OpenAI gets wrong about voice

Mira Murati's start-up presents its first AI model and aims to free voice AI from the question-and-answer model. The model processes audio, video and text in 200-millisecond chunks in parallel and aims to beat OpenAI's GPT Realtime 2 and Google's Gemini Live in terms of interaction quality. The article Thinking Machines Lab ships its first model and argues interactivity is what OpenAI gets wrong about voice appeared first on The Decoder.

5 Useful Python Scripts for Time Series Analysis

Time series data is common across finance, operations, engineering, and research. These five Python scripts cover the analysis tasks that come up repeatedly.

Sam Altman's personal investments face political scrutiny ahead of OpenAI's planned IPO

Sam Altman's personal investments face political scrutiny ahead of OpenAI's planned IPO. The article Sam Altman's personal investments face political scrutiny ahead of OpenAI's planned IPO appeared first on The Decoder.

How NVIDIA engineers and researchers build with Codex

Teams use Codex with GPT-5.5 to ship production systems and turn research ideas into runnable experiments.

What Parameter Golf taught us about AI-assisted research

Parameter Golf brought together 1,000+ participants and 2,000+ submissions to explore AI-assisted machine learning research, coding agents, quantization, and novel model design under strict constraints.

AutoScout24 scales engineering with AI-powered workflows

Learn how AutoScout24 Group uses Codex and ChatGPT to speed development cycles, improve code quality, and expand AI adoption.

Building Blocks for Foundation Model Training and Inference on AWS

The EU wants to regulate AI but needs OpenAI and Anthropic to let regulators through the door

OpenAI has offered the EU Commission direct access to its new GPT-5.5 Cyber model for security review, with talks already underway. Anthropic is proving harder to pin down: after four to five meetings on its Mythos model, regulators still don't have access. The gap highlights how dependent Europe's AI oversight remains on voluntary cooperation from the companies it aims to regulate. The article The EU wants to regulate AI but needs OpenAI and Anthropic to let regulators through the door appeared

Three things in AI to watch, according to a Nobel-winning economist

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. A few months before he was awarded the Nobel Prize in economics in 2024, Daron Acemoglu published a paper that earned him few fans in Silicon Valley. Contrary to what Big Tech…

Baidu's Ernie 5.1 cuts 94 percent of pre-training costs while competing with top models

Baidu's Ernie 5.1 uses just a third of its predecessor's parameters and reportedly cost only six percent of what comparable models require to pre-train. That's possible thanks to a "Once-For-All" approach that extracts smaller sub-models from a single training run. On the Search Arena leaderboard, Ernie 5.1 ranks 4th globally, behind two Claude Opus variants and GPT-5.5 Search. The article Baidu's Ernie 5.1 cuts 94 percent of pre-training costs while competing with top models appeared first on T

Guardrails for LLMs: Measuring AI ‘Hallucination’ and Verbosity

This article discusses how to implement an infrastructure for measuring and controlling overly verbose LLM responses.

OpenAI's DeployCo subsidiary adopts Palantir's playbook, building a moat from workflows no lab can simulate

OpenAI is building a consulting and implementation business. The "OpenAI Deployment Company," internally called DeployCo, is a majority-controlled subsidiary designed to help companies integrate AI systems into their core operations. The article OpenAI's DeployCo subsidiary adopts Palantir's playbook, building a moat from workflows no lab can simulate appeared first on The Decoder.

Lawsuit claims ChatGPT coached FSU shooter on gun operation, timing, and victim thresholds

OpenAI is facing a lawsuit over the mass shooting at Florida State University. According to the complaint, the shooter spent months talking to ChatGPT about guns and shootings. Florida's attorney general has launched a criminal investigation, saying: "If ChatGPT were a person, it would be facing charges for murder." The case adds to a growing wave of lawsuits targeting AI chatbots. The article Lawsuit claims ChatGPT coached FSU shooter on gun operation, timing, and victim thresholds appeared fir

How ChatGPT adoption broadened in early 2026

ChatGPT adoption surged in Q1 2026, with fastest growth among users over 35 and more balanced gender usage, signaling broader mainstream AI adoption.

Build an AI-Powered Learning Management System That Actually Trains People

Learn how to build an AI-powered Learning Management System from scratch using Ollama, FastAPI, and React. A step-by-step guide for beginner and intermediate developers.

AI turns patches into working exploits in 30 minutes, and the 90-day disclosure window is the casualty

Language models find security flaws faster and turn patches into working exploits in minutes. A veteran researcher says the established disclosure process needs to change. The article AI turns patches into working exploits in 30 minutes, and the 90-day disclosure window is the casualty appeared first on The Decoder.

Fostering breakthrough AI innovation through customer-back engineering

Despite years of digitization, organizations capture less than one-third of the value expected from digital investments, according to McKinsey research. That’s because most big companies begin with technological capabilities and bolt applications onto them, rather than starting with customer needs and working backward to technology solutions. Not prioritizing the customer can create fragmented solutions; disjointed…

Implementing advanced AI technologies in finance

In finance departments that have long been defined by precision and control, AI has arrived less as a neatly managed upgrade than as a quiet insurgency. Employees are already using it while leadership races to impose structure, governance, and strategy after the fact. The result is a paradox: one of the most tightly regulated functions…

Generative AI turns identity theft into an industrial-scale operation

A Bloomberg investigation shows how generative AI and autonomous agents are supercharging identity theft in the US, from social security number lookups on the darknet to deepfake driver's licenses. The article Generative AI turns identity theft into an industrial-scale operation appeared first on The Decoder.

Nvidia pumps over 40 billion dollars into AI partners so far in 2026

Nvidia has invested more than $40 billion in AI companies in 2025, cementing its role as the industry's biggest backer. The article Nvidia pumps over 40 billion dollars into AI partners so far in 2026 appeared first on The Decoder.

Import AI 456: RSI and economic growth; radical optionality for AI regulation; and a neural computer

What laws does superintelligence demand?

10 GitHub Repositories to Master FastAPI

Learn FastAPI through templates, examples, guides, auth tools, microservices, full-stack starters, and machine learning projects.

OpenAI's internal share sale minted roughly 75 multimillionaires who each cashed out the $30 million cap

In October 2025, OpenAI organized a $6.6 billion share sale for over 600 current and former employees. Around 75 of them cashed out the maximum cap of $30 million each. President Greg Brockman says he holds shares worth roughly $30 billion. The article OpenAI's internal share sale minted roughly 75 multimillionaires who each cashed out the $30 million cap appeared first on The Decoder.

OpenAI Campus Network: Student club interest form

Join the OpenAI Campus Network—connect student clubs worldwide, access AI tools, host events, and build an AI-powered campus community.

How enterprises are scaling AI

How enterprises scale AI: from early experiments to compounding impact through trust, governance, workflow design, and quality at scale.

The new AI-powered Google Finance is expanding to Europe.

This week, the new, AI-powered Google Finance is launching across Europe, with full local language support. This reimagined experience offers a suite of powerful capabil…

OpenAI launches DeployCo to help businesses build around intelligence

OpenAI launches DeployCo, a new enterprise deployment company built to help organizations bring frontier AI into production and turn it into measurable business impact.

MachinaCheck: Building a Multi-Agent CNC Manufacturability System on AMD MI300X

AI agents that hack computers and replicate themselves, and they're getting better fast

Palisade Research shows that AI agents can hack remote computers, copy themselves onto them, and form replication chains. In one year, the success rate jumped from 6 to 81 percent. The researchers expect remaining barriers to fall as models get better at hacking. The article AI agents that hack computers and replicate themselves, and they're getting better fast appeared first on The Decoder.

AI agents can now hack computers and copy themselves, and they're getting better fast

Palisade Research shows that AI agents can hack remote computers, copy themselves onto them, and form replication chains. In one year, the success rate jumped from 6 to 81 percent. The researchers expect remaining barriers to fall as models get better at hacking. The article AI agents can now hack computers and copy themselves, and they're getting better fast appeared first on The Decoder.

Anthropic and OpenAI sit down with religious leaders to seek ethical advice

Anthropic and OpenAI are turning to religious leaders for help with AI ethics. At the first "Faith-AI Covenant" roundtable in New York, representatives from both companies met with faith leaders from various religions. Critics like AI researcher Rumman Chowdhury call the talks "at best a distraction" from concrete questions about regulation and control over AI systems. The article Anthropic and OpenAI sit down with religious leaders to seek ethical advice appeared first on The Decoder.

ByteDance plans over $30 billion for AI expansion, bets big on Chinese chips

ByteDance is raising its planned AI spending for 2026 to over 200 billion yuan (roughly $30 billion), at least a 25 percent jump from earlier plans. The TikTok parent is increasingly turning to Chinese chips. Still, the figure looks modest next to the $725 billion that Google, Amazon, Microsoft, and Meta are planning to spend combined. The article ByteDance plans over $30 billion for AI expansion, bets big on Chinese chips appeared first on The Decoder.

METR says it can barely measure Claude Mythos, Palo Alto Networks warns of autonomous AI attackers

METR can barely measure Claude Mythos Preview with its current test suite. Only five out of 228 tasks cover the relevant capability range. Meanwhile, Palo Alto Networks reports that frontier models autonomously chain vulnerabilities, shrinking the time from initial access to data exfiltration to just 25 minutes. Evaluation methods are growing more slowly than the models themselves, and that may be the bigger problem. The article METR says it can barely measure Claude Mythos, Palo Alto Networks w

GPT-5.5 costs 49 to 92 percent more than its predecessor, depending on the input length

OpenAI doubled GPT-5.5's list price compared to GPT-5.4, claiming shorter responses would offset the increase. An OpenRouter analysis of real usage data tells a different story: actual costs rose 49 to 92 percent depending on input length. Anthropic hiked Opus 4.7 prices too—and with both companies eyeing IPOs, the trend is unlikely to stop. The article GPT-5.5 costs 49 to 92 percent more than its predecessor, depending on the input length appeared first on The Decoder.

Researchers may have found a way to stop AI models from intentionally playing dumb during safety evaluations

A study by researchers from the MATS program, Redwood Research, the University of Oxford, and Anthropic examines a safety problem that grows more pressing as AI systems become more capable: "sandbagging," where a model deliberately hides its true abilities and delivers work that looks adequate but is intentionally subpar. The article Researchers may have found a way to stop AI models from intentionally playing dumb during safety evaluations appeared first on The Decoder.

"OncoAgent: A Dual-Tier Multi-Agent Framework for Privacy-Preserving Oncology Clinical Decision Support"

Fields Medalist says ChatGPT 5.5 Pro delivered "PhD-level" math research in under two hours with zero human help

Fields Medalist Timothy Gowers had ChatGPT 5.5 Pro tackle open problems in number theory. The model improved an exponential bound to a polynomial one in under an hour. An MIT researcher involved calls the key idea "completely original." Gowers' takeaway: the bar for mathematical contributions is now proving something LLMs can't. The article Fields Medalist says ChatGPT 5.5 Pro delivered "PhD-level" math research in under two hours with zero human help appeared first on The Decoder.

Broadcom reportedly won't build OpenAI's custom chip unless Microsoft buys 40 percent of them

OpenAI's custom AI chip project with Broadcom has hit a funding wall. Broadcom won't finance production unless Microsoft commits to buying 40 percent of the chips, and Microsoft hasn't agreed yet. OpenAI manager Sachin Katti called the dependency "financially unattractive" in an internal message. The first phase alone costs around 18 billion dollars. The article Broadcom reportedly won't build OpenAI's custom chip unless Microsoft buys 40 percent of them appeared first on The Decoder.

Google's "Preferred Sources" feature is a free pass for more garbage in search

Google frames "Preferred Sources" as a way to bring more quality journalism into search. In practice, it shifts responsibility to a manual setting almost no one will use. That gives Google a user-choice argument for users and regulators while it keeps sidelining the open web in favor of its own AI interfaces. The article Google's "Preferred Sources" feature is a free pass for more garbage in search appeared first on The Decoder.

Pseudoscientific emotion AI is invading the workplace, an Atlantic report shows

Software that claims to read human emotions using AI is quietly becoming a fixture of everyday work life, Ellen Cushing reports in a feature for The Atlantic. The article Pseudoscientific emotion AI is invading the workplace, an Atlantic report shows appeared first on The Decoder.

Musk v. Altman week 2: OpenAI fires back, and Shivon Zilis reveals that Musk tried to poach Sam Altman

In the second week of the landmark trial between Elon Musk and OpenAI, Musk’s motivations for bringing the suit were under scrutiny. Last week, Musk took the stand, alleging that OpenAI CEO Sam Altman and president Greg Brockman had deceived him into donating $38 million to the company. He claimed that they’d promised to maintain…

AI money keeps flowing as Deepseek plans record raise and Core Automation quadruples valuation in weeks

Deepseek is planning a funding round of up to $7.35 billion, the largest ever for a Chinese AI company. Deepseek V4.1 is set to launch in June. Meanwhile, Core Automation, founded by ex-OpenAI researcher Jerry Tworek just six weeks ago, is already targeting a $4 billion valuation. The article AI money keeps flowing as Deepseek plans record raise and Core Automation quadruples valuation in weeks appeared first on The Decoder.

CyberSecQwen-4B: Why Defensive Cyber Needs Small, Specialized, Locally-Runnable Models

SoftBank reportedly slashes OpenAI-backed loan from $10 billion to $6 billion as lenders balk at private AI valuations

SoftBank has reduced a loan secured by OpenAI shares from 10 to around 6 billion dollars. Lenders are apparently reluctant to reliably assess the value of an unlisted company like OpenAI. The article SoftBank reportedly slashes OpenAI-backed loan from $10 billion to $6 billion as lenders balk at private AI valuations appeared first on The Decoder.

EMO: Pretraining mixture of experts for emergent modularity

See what happens when creative legends use AI to make ads for small businesses.

Today we're launching The Small Brief, an initiative bringing together three ad industry icons to champion a local businesses they love. Their mission is to build breakt…

Anthropic approaches $1 trillion valuation as revenue grows fivefold

According to the Financial Times, Anthropic's planned funding round is taking shape. The round aims to raise up to $50 billion, which would value the company at roughly $900 billion. The article Anthropic approaches $1 trillion valuation as revenue grows fivefold appeared first on The Decoder.

Stop Wasting Tokens: A Smarter Alternative to JSON for LLM Pipelines

If you are feeding structured data into an LLM, there is a good chance you are paying a JSON tax.

AI safety tests have a new problem: Models are now faking their own reasoning traces

Anthropic's Natural Language Autoencoders make Claude Opus 4.6's internal activations readable as plain text. Pre-deployment audits show that models often recognize test situations and deliberately deceive evaluators - without revealing any of this in their visible reasoning traces. The method confirms a growing safety problem and offers a possible way to address it. The article AI safety tests have a new problem: Models are now faking their own reasoning traces appeared first on The Decoder.