AI & Technology
Anthropic heads toward a $2 trillion IPO in November The Claude maker is set to list as one of the biggest tech offerings in history, despite still being unprofitable.
2 sources
Still developing
OpenAI Safety Employee Resigns, Sounding the Alarm on Company Culture David Robinson, who used to write the safety reports that accompanied every major model release at OpenAI, has resigned from his position and is speaking out in an editorial in The Atlantic. TechCrunch The Verge ServiceNow launches Flow, a one-day AI service desk for Slack and Teams A standalone conversational service desk needs no CMDB migration or implementation project, and targets the Fortune 500,000.
Meta open-sources the code for Muse gadgets Meta is handing out firmware and a Linux SDK so hobbyists can wire its Muse AI agent into displays, Raspberry Pis, and ESP32 boards. The Verge TechCrunch Google Questions Space Data Center Viability Google has cast doubt on the practicality of SpaceX's space data center ambitions, suggesting that Starship would need to launch 1,800 times before the concept could realistically take off. TechCrunch Oracle launches Fusion Claw, adding 25 agentic AI apps to cloud suite The Oracle acquisition of Fusion Claw shifts the company from "task assistance" to "running complete business processes," directly competing with Salesforce and ServiceNow. 247 Wall St Nvidia Unveils $4,999 DGX Spark Amid Memory Crunch Nvidia has debuted the DGX Spark, a new AI development system priced at $4,999 that features half the RAM and storage of its predecessor. The Register Apple tightens Mac full disk access to rein in AI agents After an AI agent surfaced a user's private message without a clear grant of permission, Apple says it will now require very explicit user action to hand over this sweeping permission. TechCrunch Ars Technica The Verge Cloudflare launches a Web Search API for AI agents A new beta product lets agents and applications ground their responses in live web search results from three providers, all with zero data retention.
Trump unveils the Super Intelligence Force The task force, chaired by intelligence director Jay Clayton, has 120 days to assess the risks and opportunities of AI.
AI & Technology AI & Technology Optogenetics wins the 2026 Nobel Prize in Medicine
Karl Deisseroth, Peter Hegemann and Georg Nagel are honored for the discovery of light-gated ion channels and the development of optogenetics.
1 source →
ServiceNow & ITSM ServiceNow & ITSM Customer-built AI agent on ServiceNow hits narrower pain points than the roadmap assumed
A first-hand account finds conversational AI freed the team from dashboards, record analysis, and repetitive admin work.
1 source →
ServiceNow & ITSM ServiceNow's AutoSynthData lifts a Gemma model's ITSM benchmarks by up to 7.2 points
A synthetic-data pipeline that turns model failures into training tasks measurably improves enterprise-workflow performance.
1 source →
AI & Technology
Anthropic heads toward a $2 trillion IPO in November The Claude maker is set to list as one of the biggest tech offerings in history, despite still being unprofitable.
The Claude maker is set to list as one of the biggest tech offerings in history, despite still being unprofitable.
Anthropic is expected to go public in late November 2026, with an investor roadshow slated to begin the week of November 9 and a public S-1 filing expected before the end of October. The company confidentially filed a draft registration statement with the SEC on June 1, 2026, and is targeting a valuation of roughly $2 trillion, potentially raising around $100 billion in the offering.
The listing is a major test of investor appetite for frontier AI: revenue has been growing at an exponential pace, but the company remains loss-making, and its prospectus points to plans to spend some $500 billion on cloud computing over the coming years. Anthropic is part of a 2026 wave of AI IPOs that also includes OpenAI and SpaceX, and its safety-conscious positioning is being leaned on as a selling point in a market that has grown wary of AI spending.
2 sources
The Register
Zacks
AI & Technology
Trump unveils the Super Intelligence Force The task force, chaired by intelligence director Jay Clayton, has 120 days to assess the risks and opportunities of AI.
The task force, chaired by intelligence director Jay Clayton, has 120 days to assess the risks and opportunities of AI.
President Trump announced the formation of the Super Intelligence Force on Sunday on Truth Social, calling it the federal government's vehicle to "ensure that America continues to lead the World in Super Intelligence," the term his executive order directed agencies to use instead of artificial intelligence. Director of National Intelligence Jay Clayton will chair the group, with FTC Chair Andrew Ferguson, Under Secretary of War for Research and Engineering Emil Michael, and OPM Director Scott Kupor as vice chairs, all reporting to Trump and White House Chief of Staff Susie Wiles.
The force follows the "Historic White House Accord on Super Intelligence" signed last week by the leaders of OpenAI, Anthropic, Google, Meta, Nvidia and SpaceXAI, who committed to voluntary safety standards including "four layers of controls and audits." Clayton, who has warned it would be irresponsible for the United States to "back away" from AI while China advances, is reportedly expected to deliver a report on AI's risks and opportunities within 120 days, with a charter that pledges to respond to "SI-enabled threats to our society, while preventing overregulation and regulatory capture."
2 sources
CBS News
TechCrunch
AI & Technology
Optogenetics wins the 2026 Nobel Prize in Medicine Karl Deisseroth, Peter Hegemann and Georg Nagel are honored for the discovery of light-gated ion channels and the development of optogenetics.
Karl Deisseroth, Peter Hegemann and Georg Nagel are honored for the discovery of light-gated ion channels and the development of optogenetics.
The Nobel Assembly at Karolinska Institutet awarded the 2026 prize, worth 12 million Swedish kronor shared equally, "for their discoveries concerning light-gated ion channels and optogenetics." Peter Hegemann and Georg Nagel discovered channelrhodopsin, a light-sensitive protein from the single-celled alga Chlamydomonas in which blue light opens an ion channel, and Karl Deisseroth applied it to the brain: he introduced the protein into rat nerve cells in 2005 and made the light-controlled switch work in the brains of living mice by 2007.
Optogenetics lets researchers probe how nerve cells shape memories, feelings and behavior in the living brain, establishing causal links that earlier methods could not, and is already being used to map circuits relevant to neurological and psychiatric disorders and to attempt to restore sight in people with visual impairment. Committee chair Per Svenningsson said, "Optogenetics provides opportunities for mapping the brain in a way that we could once only dream of."
1 source
Nobel Prize
AI & Technology
Cloudflare launches a Web Search API for AI agents A new beta product lets agents and applications ground their responses in live web search results from three providers, all with zero data retention.
A new beta product lets agents and applications ground their responses in live web search results from three providers, all with zero data retention.
The Web Search API, available now in beta, lets AI agents and applications search the Internet and ground their responses in live information instead of guessing URLs or relying on a model's training cutoff. At launch it supports three search providers, Ceramic.ai, Exa and Linkup, all of which support Zero Data Retention for requests made through Cloudflare and have committed to Cloudflare's verified bot crawling standards.
Because the API runs through Cloudflare's AI Gateway, search requests appear in gateway logs and are billed to AI Gateway credits at each provider's list API price with no markup, and developers can also bring their own provider API key. It is a practical step for developers building agents that need up-to-date facts, and a sign of Cloudflare positioning its AI Gateway as a hub for agentic infrastructure.
1 source
Cloudflare Developers
ServiceNow & ITSM
ServiceNow launches Flow, a one-day AI service desk for Slack and Teams A standalone conversational service desk needs no CMDB migration or implementation project, and targets the Fortune 500,000.
A standalone conversational service desk needs no CMDB migration or implementation project, and targets the Fortune 500,000.
ServiceNow introduced Flow, a standalone, AI-native service desk that works inside Slack, Microsoft Teams, and email and can go live in a single day with no infrastructure, no implementation project, and no employee training. When a request resolves correctly, users can convert that resolution into a reusable automation with one click, so repeat password resets, access requests, and policy questions are handled automatically going forward. Flow is in controlled availability now, with North American customers getting priority access, and general availability is expected for North America and EMEA in Q4 2026. Existing customers on AI-native SKUs can deploy it through consumption-based usage with no additional procurement, while net-new customers can buy it standalone with no prior ServiceNow relationship required.
Amit Zavery, ServiceNow's President, COO, and CPO, framed Flow as an investment in AI-native market opportunities, a kind of startup within ServiceNow that operates outside the company's standard release cadence. Futurum analysts read it as a preemptive move against AI-native challengers and against Freshworks and Atlassian in the small and midmarket segment, where ServiceNow's traditional platform model has been weakest, and where buyers increasingly want to reject months-long implementations outright.
3 sources
ServiceNow Newsroom
Futurum Group
Wisevoter
ServiceNow & ITSM
Customer-built AI agent on ServiceNow hits narrower pain points than the roadmap assumed A first-hand account finds conversational AI freed the team from dashboards, record analysis, and repetitive admin work.
A first-hand account finds conversational AI freed the team from dashboards, record analysis, and repetitive admin work.
A team that built an AI agent for ServiceNow reports that the real pain points it needed to solve were narrower than the company's roadmap had assumed. By experimenting with conversational AI, the team says it now spends less time navigating dashboards, analyzing records, and performing repetitive administrative work, and more time acting on what the work actually requires.
The write-up is a practitioner reality check on agentic AI on the platform: value concentrates in specific, narrow workflows rather than sweeping transformation, and it lands just as ServiceNow pushes Flow and the broader AI platform toward smaller buyers.
1 source
VentureBeat
ServiceNow & ITSM
ServiceNow's AutoSynthData lifts a Gemma model's ITSM benchmarks by up to 7.2 points A synthetic-data pipeline that turns model failures into training tasks measurably improves enterprise-workflow performance.
A synthetic-data pipeline that turns model failures into training tasks measurably improves enterprise-workflow performance.
ServiceNow AI published results for AutoSynthData, a pipeline from its CoreAI team that converts a model's failures into new synthetic training tasks for supervised fine-tuning. On a hybrid "EnterpriseOps Gym" domain, roughly 2,000 generated samples raised Gemma-4-26B-A4B-it's mean Pass@1 by 7.2 percentage points, a 35 percent relative improvement that closed 59 percent of the gap to a stronger reference model in about 18 hours. On an ITSM-domain benchmark, mean Pass@1 rose from 18.77 percent to 27.18 percent.
FourWeekMBA's analysis notes the figures are ServiceNow's own claims, based solely on supervised fine-tuning, with the company planning to test the approach with reinforcement learning; it also flags that the results come from a single post with no independent verification yet.
1 source
FourWeekMBA