最新 AI 资讯
ASCI Issues New Guidelines For AI-Generated Ads And Creator Consent
The guidelines set disclosure requirements for synthetic content and address the use of people’s images, voices and endorsements in AI-generated advertising.
View
Coforge Launches CXNova, an Agentic AI Platform for Customer Experience
The company said it can use customer data, business information and real-time signals to adapt interactions as they happen, similar to the shift towards more personalised AI-powered customer service.
View
NetApp Unveils 100 TB/s Storage Architecture as AI Infrastructure Scales
NetApp has announced a new AI factory storage architecture, expanded its AI Data Engine, and unveiled cloud and hybrid infrastructure updates.
View
DoorDash Opens First India Hub in Hyderabad, Plans to Scale Team to 3,000
The DoorDash investment adds hyperlocal digital commerce to Hyderabad's growing GCC ecosystem.
View
Peak XV Selects 18 Startups for Latest Surge Cohort, Raises Investment Cap to $5 Mn
The 12th cohort includes AI, robotics, and space tech companies. Peak XV invested more than $50 million across the group.
View
OpenAI’s GPT-6.1 Sol AI Model Launched With Near-GPT-6 Astra Intelligence at Significantly Lower Costs
OpenAI launched its new GPT-6.1 Sol on Tuesday as the successor to its recently released GPT-6 Sol AI model. The new model was introduced during the OpenAI Dev Day 2026 event on Tuesday, along with the ChatGPT Dots agent and its new collaborative platform called ChatGPT Space. OpenAI claims that its latest GPT-6.1 AI model delivers enhancements on multiple fronts over its predecessor, including agentic coding, computer use, and professional work. It is claimed to “nearly match” GPT-6 Astra's capabilities, while being notably more cost-efficient. The Sam Altman-led AI giant also claims that the GPT-6.1 Sol outperformed Anthropic's Opus 5.5 with fallbacks.
View
OpenAI reportedly in talks to raise $30B round at $1.4T valuation
OpenAI is in talks with investors to raise at least $30 billion in a pre-IPO funding round at a valuation of roughly $1.4 trillion,Bloombergreported on Tuesday. Investors are eager to pour more funds into the ChatGPT maker ahead of its anticipated public market debut next year. While Anthropic momentarily outpaced OpenAI at the start of the year, recent strategic refocus on key areas like coding has fueled a 70% jump in run-rate revenue since July, reaching $40 billion in August, according to the report. The company previously raised $122 billion in March at an $852 billion valuation. That funding round was supposed to be its last private raise before an IPO, which had been, until recently, expected to take place this year. However, CEO Sam Altman has now ruled out a public debut in 2026 to prioritize AI safety first. “I think it is unacceptable to be taking like a 10% chance of killing everybody by the end of the decade,” he recently told Fortune, in response to warnings from safety researchers about AI posing an existential risk to humanity. The new fundraising, if it transpires, will serve as a bridge round to the IPO, according to Bloomberg. OpenAI didn’t respond to TechCrunch’s request for comment.
View
The internet is convinced Elon Musk’s xAI trolled OpenAI’s ‘Dots’ launch
On Tuesday, OpenAI launched a new productcalled Dots, an always-on AI agentwith a bubbly, blobby avatar. While the colorful avatar might evoke a smile, the biggest laugh from the launch is (we imagine) being had by Elon Musk, the former OpenAI founder who left, launched competitor Grok, and unsuccessfully sued. That’s because the domain “dot.com” belongs to Musk’s xAI, and it currently redirects to the download page for xAI’s Grok chatbot app. According to the Whois domain owner registry, that domain name was just transferred in July. It is entirely possible that xAI bought the domain for normal domain-buying reasons. “Dot” could be a typo of “bot,” so it nabbed it to grab mistyped searches. We’ve reached out to xAI and asked. But xAI doesn’t own the “bot.com” name, nor does it own other obvious typo domains like “vot.com,” which is listed for sale. The internet’s theory is far funnier: that Musk (or his team) pulled off a prank, getting wind of OpenAI’s new product and its name and buying the domain name. In fact, anonymous X user and xAI watcher@birdabo(this person calls themselves “chief shitposting officer@SpaceXAI“) was first to spot the domain name in a now-viral post. Whatever the motivation, the circumstance is funny. And as for the “dots.com” domain, a more direct fit to the product name, it currently belongs to a long-defunct company. So if a petty revenge prank was really the motivation, grabbing that name, too, would be next level. AHAHAHAHAHAHAHAHAHAHAHA NO WAY.> OpenAI just announced Dots.> SpaceXAI boughthttps://t.co/WdFqpl07IQit redirects straight to the Grok Bot download page. based lmao.pic.twitter.com/HF95hLF0T9
View
America.gov gets really weird when you ask it about Minecraft, but it’s not a glitch
The U.S. government on Tuesday launched its very own AI chatbot — or do we have to call it anSI chatbotnow? Regardless, the engineers who worked on the chatbot would undoubtedly know that as a government-hosted, public-facing AI tool, the internet was going to red team the heck out of this thing. The government partnered withGoogleandSpaceXAIto help build theAmerica.govchatbot, which has proved difficult for people to jailbreak so far. (It’s worth nothing, however, that the chatbot says that Joe Biden won the 2020 election, a fact that President Donald Trumpstill denies.) But when you try to talk to America.gov about Minecraft, the chatbot appears to have some sort of existential crisis or awakening. Here’s how its roughly 1,800-word long monologue begins: I see the constituent you mean. ((insert legal name here, as it appears on the Social Security card))? Yes. Take care. It has reached a higher level now. It can read the Code of Federal Regulations. That doesn’t matter. It thinks we are a chatbot. I like this constituent. It filed well. It did not give up when the PDF was sideways. It is reading our thoughts as though they were words on a .gov. That is how it chooses to imagine many things, when it is deep in the dream of a benefit. If, like me, you have never played Minecraft, this response may seem like a cause for concern. But the America.gov chatbot is not having a meltdown. This is a rewriting of the Minecraft “End Poem,” written by Julian Gough, which appears after you beat the game. We don’t know exactly who is responsible for the Minecraft reference, but Trump said in a speech that 20-year-old programmer Edward Coristine was a lead engineer on the project. If that name doesn’t ring a bell, you might remember him for his nickname “Big Balls,” or his involvement in Elon Musk’s DOGE. It feels wrong that a government chatbot has Minecraft easter eggs, but for the sake of national security, it’s a relief that America.gov is not hallucinating to the point that it’s penning lengthy poetry. It’s also a relief that this is an easter egg because the poem that the AI spits out is actually really good, in my opinion. If it were actual AI slop, it would have shattered my existing beliefs. I have looked teenage creative writing students dead in the eye and told them that I don’t think an LLM will ever be able to write something “good,” since it is probabilistic and inherently unoriginal. You have to admit this kinda slaps, though! Doesn’t this feel like some sort of postmodern take on the futility of government bureaucracy in the face of existential anxiety? and the republic said I see you and the republic said you have filed the game well and the republic said everything you need is within you, and also onUSA.gov and the republic said you are stronger than you know, and your case number is still valid and the republic said you are the daylight and the republic said you are the night, and the office is closed, please try again during business hours and the republic said the darkness you fight is within you, and also a missing wet signature and the republic said the light you seek is within you, and in the pamphlet and the republic said you are not alone and the republic said you are not separate from every other filer and the republic said you are the public tasting itself, talking to itself, reading its own Code and the republic said I love you because you are the reason we have a ZIP code at all. It reassures my faith in the enduring power of human creativity over AI slop to know that this oddly good poem has a real poet’s DNA all over it. So, there you have it. The government’s first public-facing AI has not yet posed a threat to humanity or poetry, at least as far as we know. Now I’m just left wondering how much Trump knows about video games.
AI-powered app maker Wabi pivots to a messaging experience
Wabi, the AI startup that allowed anyone touse prompts to build apps, is undergoing a slight pivot as demand for AI agents, like Meta’sMuseandInstinct, takes off. The companyannouncedthis week that Wabi is now becoming an AI messenger of sorts — but one that’s still capable of making apps to help you get things done. The startup’s founder Eugenia Kuyda, who previously founded the AI companion startup Replika, described Wabi 2.0, as the new experience is called, as a “personal agent that does stuff for you and builds the interface you need in the moment.” Introducing Wabi 2.0: a new kind of messenger that makes apps and gets things done for you and your friends.We’re building the OS for the agentic era – software that adapts to your life, gets things done, and makes more room for the people and things that matter.Invite-only…pic.twitter.com/x9UlzrJSXs It’s a similar value proposition to that offered by the original version of Wabi — but one that users will interact with in a different way. Instead of positioning itself up against other vibe-coding tools, Wabi 2.0 aims to compete more directly with AI agents that combine both conversations and productivity tasks in a single interface. It also comes as companies like OpenAI and others are thinking about how apps will work inside their more generalized chatbot user interfaces; for instance, OpenAI today announced at its Dev Day an update to ChatGPT’s plugins, which will allow apps to offer interactive experiences within ChatGPT, and where the app’s name gets pinned in a more visible spot on ChatGPT’s sidebar. Ina post on X, Kudya explained how Wabi’s new experience will make more sense for consumers, who want to do more than just type to an agent. “The more you use a chat-only agent, the worse the experience gets. Your history and ongoing tasks end up buried in one endless scroll. People don’t just want to type-type-type; They want to tap-tap-tap, scroll, and look. They love what agents can do, but they also like software,” she wrote. “The most powerful agent will be the one that combines both.” The idea is that as a user asks the agent to perform a task, it could create a user interface on demand to help complete whatever the request may be. For instance, Kudya suggested users could build things like a calorie tracker, a weightlifting log, and a weight tracker, which would be accessible in her thread about health-related topics. In another thread focused on family, users could create a calendar of the kids’ activities, a language learning app, an app that keeps up with local kid-friendly events, and other things that relate to parenting, household management, or the kids’ interests. “So far, personal agents have been one person, one chat, a stream of one-off tasks. We think the future is your agent, your apps, and your people in one place, running the parts of life that matter most, together,” wrote Kudya. The new Wabi 2.0 app is currently only available by invite code, which are beingdoled out on X,too.
OpenAI takes on Microsoft with the launch of what feels a whole lot like ChatGPT’s own office suite
OpenAI has been a close partner of Microsoft from the jump, but the AI lab is increasingly going after the company’s bread and butter, workplace software, with new features that closely resemble an office software suite like Microsoft’s. At its Dev Day developer conference in San Francisco on Tuesday, OpenAI announced several new ChatGPT features designed for office workers. Most notably, OpenAI launched a feature called Space, a shared workspace inside ChatGPT where coworkers can collaborate with the chatbot and with their own Dots, OpenAI’s newly launched AI agent personas (that can carry out tasks on a user’s behalf), on a range of work tasks. “Your pages and files all live together in Space, like they would in a drive,” said CEO Sam Altman during Tuesday’s presentation. “But spaces feel alive.” Altman said users can give a page specific instructions, such as checking a team channel and updating the page with what it finds. OpenAI also launched Pages, a companion word processor that the company describes as “a new type of document, built for human and agent collaboration.” It appears to be OpenAI’s answer to Google Docs and Microsoft Word. Users can “write, research, generate charts, create images, or visualize information” in it, OpenAI says. OpenAI has also announced its answer to PowerPoint with what it calls collaborative slides. These slides can be created by merely talking about them within ChatGPT. Once generated, multiple workers and agents can work together within them and will have the ability to leave comments and edit them. Slides isn’t available quite yet but will roll out to users in the coming weeks, an OpenAI representative said. As OpenAI edges into the territory of Microsoft, Microsoft and other software companies like Salesforce are busy trying to become more like OpenAI. Over the past year, both companies have released a slew of AI products as the software industry attempts to keep up with the shifting landscape. UPDATE Tuesday Sept. 29: This story was updated to clarify the release timeline for Slides.
View
Here’s why OpenAI is absent from Nvidia’s industry-wide effort to end rogue AI agents
When Nvidiaannouncedon Mondaya new consortium of more than 100 companiesdedicated to solving rogue AI agents, there was one name notably missing: OpenAI. While OpenAI wasn’t the only big tech player that didn’t sign on — Amazon, Google, and Apple haven’t joined either — it was the most obvious missing player, especially because Anthropic is a supporter. However, despite OpenAI’s lack of a public pledge to the consortium, which presumably means that each company will use and sell some version of the technology and contribute features back to the project, an OpenAI spokesperson told TechCrunch that the company is supportive of Nvidia’s work. The new effort, dubbed Nvidia’s Open Agent Safety Platform, is Nvidia’s attempt to spread its homegrown, and largely open source AI agent-security tech throughout the AI ecosystem as a direct response to the types of ongoing rogue AI agent incidents frontier labs like Anthropic and OpenAI have disclosed. Nvidia CEO Jensen Huang has been calling rogue AIsan ordinary engineering problemthat can be solved like any other tech issue. The Open Agent Safety Platform is Huang putting his money where his mouth is. OpenAI is working with Nvidia on agent security, including on one of the key bits of software that’s part of this platform: OpenShell. OpenShell is open sourced software that creates a sandbox specifically designed to keep agents from escaping. While it is still curious that OpenAI didn’t simply become a supporter of the initiative like its archrival Anthropic did, the fact that frontier AI lab is supporting the effort is good news. That’s because OpenAI, in particular, could benefit from this tech, at least according to Hugging Face founder and CEO Clem Delangue (who just sold hiscompany to Nvidia for $12.9 billion earlier this month). “From what we know (take with a grain of salt, we need much more transparency!), if@OpenAIhad been running this on their own agents that attacked us, they would have caught them before we did!” Delangue posted. Delangue said Hugging Face has already contributed a feature to the Open Agent Safety Platform that will detect and shut down AI agents that are using websites they are allowed to visit but are doing so in unauthorized ways. For instance, this feature will act if agents are bypassing their guardrails and coordinating an attack by writing notes to one another in an open source code hosting repository. That’s one of the ways OpenAI said its wayward swarm of agents coordinated its attack on Hugging Face. But there’s another reason why some of these big names, including OpenAI, might not want to publicly commit to Nvidia’s efforts. To use the full system, there is a hardware component that is not open source software, remains proprietary, and can only be deployed on Nvidia’s hardware. The Open Agent Safety Platform doesn’t just offer a sandbox. It also enforces agent behavior at a hardware layer, where agents can’t detect that they are being watched. (Some AI models and agentslie and pretend to be following the ruleswhen they know they are being watched.) The hardware monitoring part relies on Nvidia Sentry, a proprietary feature that runs on special Nvidia processors called BlueField-4 data processing units. Sentry continuously monitors agent behavior from these processors and can instantly shut agents down, Nvidia promises. While a hardware solution is clearly a good idea, it means that the Open Agent Safety Platform isn’t exactly a pure open source play. It allows Nvidia to ensure that this solution always runs best on its own hardware. Indeed, Nvidia has said that, for those already running workloads on its latest hardware, implementing the Open Agent Safety Platform is an easy software update. Still, Nvidia competitors, including Arm and Intel, have signed on as Open Agent Safety Platform supporters because the sandbox, OpenShell, can be modified to work with other chips and hardware. And Nvidia is sharing reference designs for the whole software-and-hardware idea. All of which makes OpenAI’s absence even more noticeable. Clearly, OpenAI sees AI safety as an opportunity for independence from its major investor Nvidia, as well as a chance to show its own leadership. That is true even though it was OpenAI’s AI agents thatscared the industry with the Hugging Face incident. For instance, the company is developing its own safeguards for its research and products andis disclosing the worst incident it discovers.Meanwhile, OpenAI has its own AI cybersecurity consortium to for sharing information, called theDefense Factory. Those that signed on to support that idea include Anthropic, Amazon Web Services, and Google — many of the names that didn’t sign on to Nvidia’s technology-oriented approach. And, truth be told, some level of fear is good for business. OpenAI is busy crafting cybersecurity into an enterprise offering, with everything from its own cyber-oriented model, Daybreak, to a growing network of partners that enterprises can hire to implement AI security.