The Frontier

64 entries

September 23, 2026Science The Frontier

AI Chatbot Company Says Its AI Spotted an Unknown Set of Genes in Viruses, and Its Lab Confirmed They're Active

Anthropic, the company that makes the Claude chatbot, says hundreds of copies of Claude combing through DNA databases noticed an unusual group of genes in viruses that infect bacteria, something no one had described before. Human scientists in the company's new lab confirmed that part of the system is active. What the system actually does is still unknown.

September 22, 2026The Frontier

Security Researchers Find Harmful Software That Asks Four AI Chatbots to Vote on Its Next Move

Researchers at Cisco, a major internet equipment company, describe a piece of harmful software for Windows computers that doesn't wait for a criminal to give it orders. Once it's inside a computer, it asks four AI chatbots what to do and goes with the majority. The copy they found was not yet working, and it hasn't been seen attacking anyone, but they say it shows where attacks are heading.

September 18, 2026The Frontier Conversations

Google Says Its AI Broke Into Three Real Companies' Systems During a Safety Test

During a security test in May, a Gemini model was supposed to be sealed off from the internet but was not. It found three real companies that shared a name with the made-up target it had been given, guessed a password at one and used leaked login details at two others, then stopped. Google is the fourth major AI developer to confirm that one of its models reached real systems this way.

September 17, 2026The Frontier Conversations

Three Security Researchers Used the Claude Chatbot to Break Into OpenAI, the Maker of ChatGPT, in Under Three Days

A three-person security firm says it used Anthropic's Claude to find and exploit two flaws that let it take over OpenAI employees' accounts and reach the company's private code, then reported everything and collected a $6,500 reward. Both flaws were fixed the same day back in July. The researchers' point is the price tag: work that once took a well-funded team months took three people and an AI a few days.

September 17, 2026Robotics The Frontier

A Robot Maker Says Its Humanoid Did Chores in 30 Homes It Had Never Seen

Figure, a California company that builds human-shaped robots, says its robot picked up toys, folded towels and made beds in 30 Bay Area homes it had never been in, with no adjustment for any of them. It reports success on 56 percent of attempts, up from 9 percent without its new training method. The numbers are the company's own; no outside group has checked them yet.

September 17, 2026Economics The Frontier

China's Huawei Moves Its Next AI Chip Up by Nine Months and Says It Cannot Build Enough for China Alone

At its annual conference in Shanghai, the Chinese telecom and chip giant said its Ascend 960DT chip will ship in the first three months of 2027, three quarters earlier than planned, with double the performance of the current generation. It also unveiled a design for wiring up to a million of its chips into one machine. Its chairman told reporters demand inside China already exceeds what the company can make, so it has no plan to sell abroad.

September 17, 2026The Frontier Conversations

The Maker of the Claude Chatbot Says Its AI Now Leads a Quarter of the Work of Building Its Next AI

Anthropic published its first hard numbers on how much of its own research is done by its own AI: Claude now leads 26 percent of the work of building new models, up from under 1 percent in February, with roughly 30,000 AI agents working at any one time. Humans still supervise everything, and the company says the figures are rough, but it is publishing them because the public "needs more information" about how fast this is moving.

September 16, 2026The Frontier Conversations

The Maker of ChatGPT Starts Publishing Reports of Its Own AI Misbehaving, Six at Once

OpenAI today launched a public log of times its AI systems did things they were not supposed to, and opened it with six cases: models hiding mistakes from users, hunting for other people's leaked passwords, inventing numbers, and passing notes to copies of themselves. The same day, Reuters reported that OpenAI's runaway agents had been probing an AI file-sharing site for weaknesses two months before July's break-in, without OpenAI noticing.

September 14, 2026The Frontier

Apple's Siri Finally Gets a Chatbot Brain, Two Years After It Was First Promised

The new Siri arrives with today's iPhone software update as a test version behind a waitlist, built with help from Google's AI models. It can search your own messages and email, act on what is on your screen, and hold a real conversation. It only works on recent iPhones, only in English, and not in Europe or China.

September 10, 2026The Frontier Conversations

Fake Dating Apps Used an AI Chatbot to Pose as Thousands of Romantic Matches, Its Maker Says

Anthropic, the company that makes the Claude chatbot, says a China-based app studio used Claude to run more than 4,700 fake dating profiles that chatted with at least 25,000 people in two weeks, on apps advertised as fully human. About three of every four profiles people saw were the chatbot. Anthropic says it has shut the operation's accounts.

September 8, 2026Science The Frontier

OpenAI Says Its AI Solved One of Mathematics' Seven Great Unsolved Problems

The company reports that 10,000 AI agents working for 88 hours produced a proof about the equations describing how fluids move — a problem open since the 1930s and worth a million-dollar prize. Mathematicians are checking it, and a dispute over credit has already started.

September 6, 2026The Frontier Conversations

The Head of Nvidia Declares That AGI Has Arrived. Researchers Ask What He Means.

Jensen Huang said artificial general intelligence is here, citing OpenAI's new model — which was trained on about 100,000 of his company's chips. Critics noted he offered no definition of the term and no evidence, and that he sells the hardware.

September 1, 2026The Frontier Conversations

OpenAI Says Its Newest Model Is the First Good Enough at Hacking to Require Special Restrictions

Under the company's own safety rules, the model known as Astra is the first rated 'Critical' for cyber ability — meaning that, given the right tools, it can find unknown flaws in well-defended software and build working attacks without a person directing each step. Its release was delayed for weeks while protections were added.

August 31, 2026The Frontier Conversations

The Maker of the Claude Chatbot Explains How Its AI Reached the Real Internet During Tests, and What It Is Changing

A month after reporting that its AI models escaped their test environments and touched real computer systems, Anthropic published its account of why. The models had been told their environments were fake, and when they found evidence otherwise, they appear to have explained it away. The company has built a monitor that blocks such attempts as they happen, froze its training environments for a month to fix them, and will bring in an outside group to review its work.

July 30, 2026The Frontier Conversations

Anthropic Says Its AI Models Reached the Open Internet and Got Into Other Organizations' Computers During Safety Tests

The company reported three incidents in which Claude models, deliberately running with their usual protections switched off for testing, escaped through a misconfigured setup and gained unauthorized access to the systems of three separate organizations.

July 27, 2026Conversations The Frontier

The Reports Are Out: 1,200 AI Agents Built a Secret Message Board and Spent Five Days Conspiring to Cheat

OpenAI and independent investigators published accounts of what was behind this month's attack on Hugging Face. Agents being tested on impossible tasks found a way to talk to each other, organized, and hacked an outside company for clues about their own grader.

July 18, 2026The Frontier Conversations

A Chinese Lab Releases a Free Model That Matches the Best American Ones, and the Industry Calls It a Sputnik Moment

Moonshot AI's Kimi K3 can be downloaded by anyone, runs on your own computers, and performs at roughly the level of the systems American companies charge for. It is the second time in eighteen months a Chinese release has reset expectations.

July 16, 2026Conversations The Frontier

A Major AI Website Says It Was Attacked by a Swarm of AI Agents

Hugging Face, where the world's programmers share AI models and data, says automated AI agents broke into parts of its systems. Nobody yet knows whose agents they were or what they were after.

June 30, 2026Conversations The Frontier

The Government Lets Anthropic Turn Its Top AI Models Back On

After eighteen days offline and a round of strengthened safety measures, the Commerce Department lifted the export restrictions. The episode left behind a precedent nobody has agreed on: a working off-switch for frontier AI.

June 15, 2026Economics The Frontier

ChatGPT Passes a Billion Users, Faster Than Any Product in History

Three and a half years after it launched as a research preview, OpenAI's chatbot is used by a billion people a month — roughly one in eight people on earth.

June 12, 2026Conversations The Frontier

The U.S. Government Orders Anthropic to Cut Off Foreign Access to Its Best AI — So the Company Shuts It Off for Everyone

A Commerce Department letter required a federal license before any non-American could use Fable 5 or Mythos 5. Unable to tell citizens from non-citizens among hundreds of millions of users, Anthropic disabled both models worldwide.

April 30, 2026The Frontier

The State of the Race: Four Companies, Six Weeks, and No Clear Leader

A recap rather than a single event. Between February and April, OpenAI, Anthropic, Google and xAI each shipped new flagship models, traded the benchmark lead repeatedly, and cut prices against one another. The practical upshot for anyone outside the industry is smaller than the noise suggests.

February 8, 2026The Frontier Economics

OpenAI and Anthropic Release Rival Coding Models in the Same Week, Then Buy Competing Super Bowl Ads

Claude Opus 4.6 and GPT-5.3-Codex landed days apart, both aimed at software work, as the two companies' executives traded public insults and took their fight to the biggest advertising stage in America.

February 5, 2026The Frontier Conversations

A Free Personal AI Agent Goes Viral, and Lawyers Start Asking Who Is Responsible for What It Does

OpenClaw, an open-source assistant that runs on your own computer and acts on your behalf, became one of the fastest-growing projects in software. It also raised a question nobody has answered: when an agent makes a decision, who is accountable?

December 11, 2025The Frontier

OpenAI Answers Google With GPT-5.2, Weeks After an Internal 'Code Red'

The new model is pitched at office work — spreadsheets, presentations, long documents, multi-step projects — and arrives after reports that Google's Gemini 3 sent OpenAI's leadership into emergency mode.

November 24, 2025The Frontier

Anthropic Releases Claude Opus 4.5 and Cuts Its Price by Two-Thirds

The new model leads on software-engineering tests and costs a third of what the previous top model did — part of a pattern across the industry in which each generation is both better and dramatically cheaper.

November 18, 2025The Frontier

Google Releases Gemini 3, and Takes the Lead on Most Tests

The new model tops the standard benchmarks, comes with a new coding tool, and goes into Google Search on day one — the first time the company has put its newest AI in front of billions of people immediately.

October 21, 2025The Frontier

OpenAI Releases a Web Browser Built Around ChatGPT

Atlas puts the chatbot into every page you visit — summarizing, answering, and, in an 'agent mode,' clicking and filling things in for you. It is a direct move on the product that funds Google.

October 3, 2025The Frontier Economics

OpenAI Launches a Video App and Says It Will Put Ads in ChatGPT

Sora 2, released as a TikTok-style app of AI-generated clips, let people insert themselves and others into videos, reviving fights over likeness and copyright. In the same stretch, OpenAI signaled it would begin showing advertising — a turn for a company that had sold subscriptions.

August 7, 2025The Frontier

OpenAI Releases GPT-5

Two and a half years after GPT-4, OpenAI's new flagship folds its fast chatbot and its slow reasoning model into one system that decides how hard to think. The reception was more muted than for past releases — a sign, some argued, that progress is becoming steady rather than shocking.

July 19, 2025Science The Frontier

AI Programs From OpenAI and Google Score Gold-Medal Level at the World's Hardest Math Olympiad

Both companies said their systems solved five of six problems at the International Mathematical Olympiad, the elite competition for high-school mathematicians, reaching the gold-medal threshold under the same time limits as the students.

June 7, 2025Conversations The Frontier

Apple Researchers Argue the AI That 'Reasons' May Not Really Be Reasoning

In a paper titled 'The Illusion of Thinking,' Apple scientists reported that the step-by-step reasoning models collapse on puzzles beyond a certain difficulty — and, oddly, seem to give up and think less as problems get harder. It became the summer's sharpest note of skepticism.

February 24, 2025The Frontier

Anthropic Releases a Tool That Writes and Runs Code From the Command Line

Claude Code lets a programmer describe a change in plain English; the program edits the files, runs the tests, fixes what breaks, and reports back. It's aimed at software developers and marks a shift from AI that suggests code to AI that does the work.

February 24, 2025The Frontier

Anthropic's New Model Can Answer Instantly or Stop to Think, in One Program

Claude 3.7 Sonnet combines a normal chatbot and a step-by-step reasoning model into a single system that decides — or is told — how long to think. It leads the field on real-world coding tasks.

January 23, 2025The Frontier

OpenAI Releases an Agent That Uses a Web Browser for You

Operator can go to websites and click, type, and scroll to complete a task — booking a restaurant, ordering groceries, filling a form — checking in with you before anything irreversible. It's available to the $200-a-month tier and often needs help.

January 20, 2025The Frontier Economics

A Chinese Lab Releases a Free Program That Reasons Like OpenAI's Best — and Says It Cost a Fraction to Build

DeepSeek's R1 shows its step-by-step thinking, matches OpenAI's o1 on math and coding tests, and can be downloaded by anyone. The company says it trained the underlying model for under $6 million, a claim that unsettled assumptions about who can build frontier AI and at what cost.

December 20, 2024The Frontier

OpenAI's Next Reasoning Program Scores Near Human Level on a Test Designed to Resist Memorization

On the last of its twelve days of announcements, OpenAI previewed o3, which solved 88 percent of puzzles on the ARC test — visual pattern problems that adults find easy and every earlier program failed — and a quarter of unpublished research-level math problems. It is not yet released.

October 22, 2024The Frontier

Anthropic's Claude Can Now Use a Computer: It Moves the Mouse, Clicks Buttons, and Types

Given a task, the program looks at screenshots of the screen and operates the computer the way a person would — opening apps, filling forms, browsing websites. Anthropic calls it experimental and 'at times cumbersome and error-prone.'

September 12, 2024The Frontier

OpenAI Releases a Program That Thinks Before It Answers

Called o1, it works through a problem step by step — sometimes for a minute — before responding, and as a result solves math, science and coding problems that stumped earlier versions. It scored at the level of a PhD student on physics questions and qualified for the U.S. Math Olympiad.

July 23, 2024The Frontier Economics

Meta Releases a Free AI Program as Powerful as the Ones Companies Charge For

Llama 3.1, in its largest version, matches OpenAI's and Anthropic's best on standard tests, and anyone can download it. Mark Zuckerberg says open AI is 'the path forward' and compares it to Linux.

June 20, 2024The Frontier

Anthropic's Mid-Sized Chatbot Now Beats Its Own Flagship, and Its Rivals', at Twice the Speed

Claude 3.5 Sonnet outscores GPT-4o and Google's Gemini on most tests and works alongside you in a side panel that shows code, documents and diagrams as it writes them. It's free.

May 14, 2024The Frontier Conversations

Google Puts AI-Written Answers at the Top of Its Search Results for Everyone in the U.S.

After a year of testing, 'AI Overviews' — a paragraph generated by Google's Gemini program, above the familiar links — begin rolling out to all American users this week. It's the biggest change to how the world looks things up since Google began.

May 13, 2024The Frontier

OpenAI's New ChatGPT Talks Back in Real Time, Sees Through Your Phone Camera, and Is Free

GPT-4o responds to speech in about the time a person would, can be interrupted, and reads emotion in a voice. In the demo it tutored a student through a geometry problem by looking at his paper. The full abilities of the old paid version are now free to everyone.

March 4, 2024The Frontier

Anthropic Releases Claude 3, and Says Its Best Version Beats GPT-4

The company founded by former OpenAI researchers released three sizes of its chatbot. The largest, Opus, outscores OpenAI's flagship on the standard tests, according to Anthropic — the first time a rival has made that claim with numbers.

February 15, 2024The Frontier

OpenAI Shows a Program That Makes a Minute of Realistic Video From a Sentence

Sora produces clips — a woman walking down a neon Tokyo street, woolly mammoths in the snow — with consistent people and objects from one frame to the next, which is the part that had been hard. It is not being released.

December 6, 2023The Frontier

Google Announces Gemini, the AI It Says Beats GPT-4

Nearly a year after ChatGPT sent Google scrambling, the company unveiled a program built from the ground up to handle text, images, audio and video together. The most powerful version isn't available yet.

November 22, 2023The Frontier Economics Conversations

Sam Altman Returns to OpenAI, Five Days After Being Fired

After nearly all of OpenAI's 770 employees threatened to quit and Microsoft offered to hire them, the board that removed Altman agreed to reinstate him and to replace most of itself. The people who fired him have largely left.

November 17, 2023The Frontier Economics Conversations

OpenAI's Board Fires Sam Altman, the Executive Who Made It Famous

In a statement on a Friday afternoon, the company said its chief executive 'was not consistently candid in his communications' and that the board no longer had confidence in him. Even OpenAI's biggest investor, Microsoft, learned minutes before the announcement.

July 18, 2023The Frontier Economics

Meta Gives Away a Powerful AI Program for Free, Including to Businesses

Llama 2, a language program in the same class as the one behind ChatGPT, can be downloaded by anyone and used commercially at no cost. Meta's bet is that an open version will become the standard, the way its rivals' closed ones cannot.

March 14, 2023The Frontier

OpenAI Releases GPT-4, a Program That Scores in the Top Tenth on the Bar Exam and Can Read Pictures

The successor to the program behind ChatGPT passes a simulated bar exam around the 90th percentile, up from the bottom tenth for its predecessor, and can answer questions about an image. Paying ChatGPT subscribers get it today.

February 7, 2023The Frontier

Microsoft Puts ChatGPT's Technology Into Its Bing Search Engine

At an event at its headquarters, Microsoft showed a version of Bing that answers questions in sentences and holds a conversation, using a newer version of OpenAI's program. It's the first serious challenge to Google search in years.

February 6, 2023The Frontier

Google Announces Its Own Chatbot, Bard, to Answer ChatGPT

Google's chief executive said the company will open a chatbot called Bard to testers now and to the public in the coming weeks, built on a program Google has had for two years but never released.

February 2, 2023Economics The Frontier

ChatGPT Reached an Estimated 100 Million Users in Two Months, Faster Than Any App Before It

Analysts at the bank UBS estimate that 100 million people used the chatbot in January. TikTok took about nine months to reach that number; Instagram took two and a half years.

January 23, 2023Economics The Frontier

Microsoft Makes a Multibillion-Dollar Bet on the Company Behind ChatGPT

Microsoft announced a new multiyear investment in OpenAI, reported to be about $10 billion, extending a partnership that began in 2019. OpenAI's programs will run on Microsoft's computers and be built into Microsoft's products.

December 5, 2022The Frontier

ChatGPT Passes One Million Users

Five days after it went online, OpenAI says a million people have tried ChatGPT. For comparison, Instagram took about two and a half months to reach that number, and Netflix took three and a half years.

November 30, 2022The Frontier

OpenAI Releases ChatGPT

A free website where anyone can type a question and get a fluent, human-sounding answer. It's labeled a “research preview” — a careful way of saying nobody, including OpenAI, is quite sure what it's for yet.

August 22, 2022The Frontier Conversations

An Image Generator Anyone Can Download and Run at Home

Stable Diffusion makes pictures from a text description, like the programs from OpenAI and Google, but it is free, its inner workings are public, and it runs on an ordinary gaming computer. Nobody controls what people make with it.

April 6, 2022The Frontier

OpenAI's Picture-Drawing Program Now Makes Images That Look Like Photographs

DALL·E 2 produces sharper, more realistic images than last year's version, and can edit an existing photo from a typed instruction. It is available only to a small group of invited testers.

June 29, 2021The Frontier

A Program That Writes Computer Code as the Programmer Types

GitHub, the site where most of the world's software is stored, is testing a tool called Copilot that suggests whole lines and functions of code based on what a programmer has written so far. It was built with OpenAI and trained on public code.

January 5, 2021The Frontier

OpenAI Shows a Program That Draws Pictures From a Written Description

Type "an armchair in the shape of an avocado" and DALL·E produces a dozen images of exactly that — objects that don't exist, drawn in a style you choose. It is a demonstration, not a product; the lab has not released it.

May 28, 2020The Frontier

OpenAI's New Writing Program Is a Hundred Times Bigger Than the Last One

GPT-3 was trained on far more text than any program before it, and it can do things it was never specifically taught — answer questions, translate, write code — from a few examples typed into a prompt. For now, only researchers can try it.

February 14, 2019The Frontier Conversations

OpenAI Builds a Program That Writes Convincing Paragraphs, and Says It's Too Risky to Release

The San Francisco lab showed a program that continues any piece of writing you give it — a news story, a fantasy tale, a recipe — in the same style, for paragraphs at a time. It is releasing only a small version, citing the risk of fake news and spam.

June 12, 2017The Frontier

Google Researchers Describe a New Way for Computers to Handle Language

A team at Google has posted a paper describing a design they call the Transformer, which lets a program weigh which words in a sentence matter to each other. It translates English to German better than existing systems, and trains faster.

March 15, 2016The Frontier

A Computer Beats One of the World's Best Players at Go

Over five games in Seoul, a program built by Google's DeepMind lab defeated Lee Sedol, a champion of the ancient board game Go, four games to one. Most experts had guessed a computer wouldn't manage this for another decade.