allnewscastallnewscast
Breaking News
Gillibrand and Welch Set Fall Senate Hearing Schedule For Tariff Repeal Bill— Politics & ElectionsFederal Judge Says Trump Team Defied Mail-Voting Injunction With New USPS Rule— Politics & Electionsnews1— Politics & ElectionsGillibrand and Welch Set Fall Senate Hearing Schedule For Tariff Repeal Bill— Politics & ElectionsFederal Judge Says Trump Team Defied Mail-Voting Injunction With New USPS Rule— Politics & Electionsnews1— Politics & Elections
AI & Tech

Broadcom, Teradata and startup funding reshape AInews in early September

Nic Reeve8 min read
Broadcom, Teradata and startup funding reshape AInews in early September
Broadcom’s chip plans and startup funding headline AInews week of September 4

On the week of September 4, the AInews cycle was dominated by Broadcom’s aggressive artificial intelligence chip forecasts, Teradata’s push to embed AI agents into enterprise analytics, and a $550 million funding round that lifted startup Wonderful to a $5 billion valuation.

What did Broadcom announce about its AI chip business this week?

Broadcom projected rapid growth for its artificial intelligence chip revenue over the next two years and highlighted large infrastructure orders tied to custom accelerators for major cloud and AI customers, underscoring its ambition to challenge Nvidia’s leadership in data center silicon.

Broadcom’s latest outlook came during an investor call held in early September, in which chief executive Hock Tan laid out a series of aggressive targets for the company’s data center AI revenue.

  • According to The Star, Broadcom told investors its AI chip revenue could reach about US$115 billion in fiscal 2027 and around US$230 billion in fiscal 2028, reflecting expected demand from hyperscale cloud operators and AI model builders.
  • In the same report, Broadcom said AI chip revenue alone was projected at US$21.7 billion in the fourth quarter of its current fiscal year.
  • Earlier guidance described AI semiconductor revenue of US$6.2 billion in the fourth quarter of fiscal 2025, up 66% year-on-year, illustrating how quickly the company has been revising its forecasts upward.
  • Reuters reported in a September 4 earnings-related story that Broadcom had secured more than US$10 billion in AI infrastructure orders from a new customer for custom accelerators, helping push fiscal 2026 AI revenue growth “significantly” higher than 2025.

Those orders relate to what Broadcom describes as customized AI "XPU" accelerators, designed for large-scale training and inference workloads inside cloud data centers. The company already counts three major customers for these chips, and the newly converted client brings the total to four.

In the call covered by financial news desks, Tan said AI semiconductor revenue had grown for ten consecutive quarters, reaching US$5.2 billion in the third fiscal quarter of 2025, a 63% year-on-year increase. He also confirmed he intends to lead Broadcom through at least 2030, signalling continuity as the firm bets its future on AI infrastructure.

How is the market reacting to Broadcom’s AI strategy and earnings?

Analyst commentary during the week described a gap between Broadcom’s strong reported AI numbers and investor sentiment, with shares pressured despite revenue beats as markets digested ambitious longer-term guidance and spending plans around custom accelerators and data center networking.

An AI-focused market newsletter published on September 4 noted that Broadcom had beaten Wall Street expectations on its latest earnings report but that the stock fell about 5% after the release, continuing a pattern of post-earnings sell-offs despite repeated outperformance.

  • The newsletter reported that this pattern had erased roughly US$500 billion of Broadcom’s market value since June as investors questioned how sustainable the current trajectory is, given heavy capital needs and competitive pressure from Nvidia and other chipmakers.
  • Commentary cited Broadcom’s reliance on a small number of very large cloud and AI customers for its custom accelerators as both a strength, in terms of visibility, and a risk if any major client shifts strategy or adopts alternative hardware.

Despite the mixed stock reaction, both Reuters and regional business coverage highlighted that Broadcom sees AI-related chips as the primary engine of its growth over the rest of the decade, expecting demand for accelerators, networking silicon and related infrastructure to expand as more enterprises adopt generative and agentic AI systems.

What new AI product did Teradata launch for enterprises?

Teradata launched an enterprise-grade Data Analyst Agent in the Amazon Web Services Marketplace, offering AI-assisted, conversational analytics for customers that already run workloads on AWS and want to query complex data through natural language rather than traditional business intelligence interfaces.

Teradata, best known for its data warehousing and analytics platforms, announced the availability of its Data Analyst Agent in late July with follow-on coverage in early August, positioning the tool as part of its broader AI services strategy.

  • According to Teradata’s press release from July 30, the Data Analyst Agent brings AI-assisted conversational analytics into existing AWS environments and is sold through the AWS Marketplace under the Teradata AI Services label.
  • The agent is designed to understand natural language questions from business users, translate them into analytic queries across Teradata systems, and return explanations, charts or summaries without requiring SQL expertise.
  • Futurum Group’s analysis of Teradata’s second-quarter 2026 results described the product as a key part of a “hybrid AI strategy” aimed at embedding AI into both on-premises and cloud deployments, which the firm said is gaining traction with large enterprises.

Teradata’s second-quarter numbers underline why the company is leaning into AI-assisted analytics. Futurum Group reported that revenue for the quarter came in at US$410 million, slightly above consensus estimates and flat year-on-year, while profitability improved. A Yahoo Finance summary of the same period noted net income of US$46 million and a higher full-year earnings-per-share outlook.

Yahoo’s report linked those upgrades partly to the launch of the Data Analyst Agent in the AWS Marketplace, arguing that cost control, recurring cloud revenue and new AI-led services are reshaping Teradata’s business mix and appeal to investors that want exposure to enterprise AI adoption.

How did Wonderful’s latest funding round reshape the AI startup landscape?

Israeli-Dutch startup Wonderful closed a US$550 million Series C round at a US$5 billion valuation, more than doubling its valuation in under six months and signalling strong investor appetite for companies building operating systems and orchestration tools for enterprise AI agents.

The Series C round was announced on September 2 and drew coverage from Reuters, TechCrunch, Business Wire and Bloomberg, each describing slightly different angles on the same transaction.

  • Reuters reported that Wonderful’s valuation had risen from roughly US$2 billion to US$5 billion in less than six months as demand grew for its AI operating system among enterprises.
  • TechCrunch wrote that Wonderful raised US$550 million, with Insight Partners again leading the round and existing backers Index Ventures, IVP, Vine Ventures, 9Yards and Bessemer Venture Partners participating.
  • Business Wire’s release confirmed the US$5 billion valuation and listed Salesforce as a new investor, joining the prior venture firms.
  • Bloomberg’s coverage described Wonderful as building software that coordinates AI agents across an enterprise, casting the company as part of a trend toward “agentic” AI systems that perform business tasks rather than just answer questions.
  • Yahoo Finance’s technology section similarly framed Wonderful’s product as an “AI OS” that helps large companies become “AI-native” by connecting different models and agents to workflows like customer support and internal operations.

This latest round follows a Series B announced in March, when Wonderful raised US$150 million led by Insight Partners. TechFundingNews reported at the time that the company’s tools were aimed at increasing deployment rates for AI pilot projects, noting that many proofs-of-concept never reach production.

With fresh capital, Wonderful says it plans to expand engineering teams in Israel and the Netherlands, invest in security and governance features for its operating system, and grow its global go-to-market presence with partners like Salesforce to reach more large enterprises.

Who is affected by these AI business moves and what changes next?

The week’s developments affect cloud providers, enterprises and investors: Broadcom’s forecasts reflect rising hardware demand from hyperscale platforms, Teradata’s agent targets analysts and business users, and Wonderful’s funding highlights a growing ecosystem of companies that help large organizations run AI agents at scale.

For cloud and AI infrastructure buyers, Broadcom’s multi-year guidance suggests expanding choices in accelerator hardware and networking, which could influence pricing and availability for training and inference capacity as demand for large language models and enterprise agents grows.

  • Cloud providers and major AI labs appear central to Broadcom’s plans, with Reuters reporting at least US$10 billion in custom infrastructure orders from a newly signed customer presumed to be a leading AI platform.
  • Enterprises making long-term commitments to specific accelerator stacks may benefit from competition between Broadcom and Nvidia but face lock-in risks if each vendor’s custom chips require tailored software tooling.

Inside enterprises, Teradata’s Data Analyst Agent and Wonderful’s AI operating system target different layers of AI adoption. Teradata is focused on making existing analytic environments easier to query through conversational interfaces. Wonderful concentrates on orchestrating multiple agents and models to handle tasks across business functions.

  • Analysts and business managers who rely on Teradata’s platforms could see shorter turnaround times for data requests once natural language interfaces become common in their workflows.
  • Wonderful’s customers, many of them large organizations with fragmented data and AI pilots, gain a central system to coordinate customer support agents, back-office automation and other AI tools.
  • Investors exposed to these companies face different risk profiles: Broadcom and Teradata are established listed firms, while Wonderful’s backers are betting on a fast-growing startup in a still-evolving category.

Over the coming months, key milestones to watch include Broadcom’s ability to convert guidance into realized shipments and margins, Teradata’s success in monetizing AI agents within its installed base, and Wonderful’s progress in turning its expanded war chest into sustained revenue growth rather than just valuation headlines.

Sources

  1. 1.moomoo.com
  2. 2.techcrunch.com
  3. 3.thestar.com.my
  4. 4.reuters.com
  5. 5.teradata.com
  6. 6.businesswire.com
  7. 7.finance.yahoo.com
  8. 8.teradata.com
  9. 9.aibusinessweekly.net
  10. 10.bloomberg.com
  11. 11.bloomberg.com
  12. 12.reuters.com
  13. 13.techfundingnews.com
  14. 14.futurumgroup.com
  15. 15.finance.yahoo.com

Read more →

Related Articles

Google adds xAI’s Grok 4.6 to its enterprise agent marketplace
AI & Tech

Google adds xAI’s Grok 4.6 to its enterprise agent marketplace

Google’s enterprise AI platform has added support for xAI’s Grok 4.6, expanding the model choices available to business users building agents and automations. The new listing appears in Google’s Gemini Enterprise Agent Platform documentation and places Grok 4.6 in the platform’s Model Garden, where developers can access third-party models alongside Google’s own offerings. According to xAI and Google documentation published this week, Grok 4.6 is positioned as xAI’s most capable model for coding, agentic tasks and knowledge work. The model is described as being built for long-running agents and more ambitious interactive and visual work, with a 500,000-token context window and configurable reasoning levels labeled low, medium, high and xhigh. The addition matters because enterprise teams increasingly want a single environment where they can compare and deploy multiple frontier models without rewriting their entire workflow. By making Grok 4.6 available inside Google’s enterprise agent stack, Google is giving customers another option for tasks that may benefit from longer context handling, multi-step reasoning and tool use. The model is also surfaced with a dedicated publisher-style entry, indicating that it can be selected and managed through the platform’s standard model browsing interface. xAI’s own release materials say Grok 4.6 is available through the xAI API and partner services, with pricing set at $2 per million input tokens, $0.50 per million cached input tokens and $6 per million output tokens for prompts under 200,000 tokens. For larger prompts of 200,000 tokens or more, xAI says pricing rises to $4 per million input tokens, $1 per million cached input tokens and $12 per million output tokens. Google’s documentation mirrors the availability, listing Grok 4.6 in preview inside Model Garden. The model’s arrival on Google’s platform follows a broader rollout that xAI announced earlier in August. In its release notes, xAI said Grok 4.6 is intended for coding, agentic tasks and knowledge work, and that it supports text and image input with text-only output. The company also says the model has no stated text output limit and includes tools such as function calling, web search, X search and code execution. For enterprise customers, the practical appeal is straightforward: Grok 4.6 is being offered as a high-capacity model for jobs that stretch over long sessions, such as software development, research synthesis and multi-step workflow automation. The 500,000-token window gives the model room to hold far more context than many standard systems, while the reasoning controls allow users to adjust how aggressively the model thinks before responding. Google has not said Grok 4.6 will replace any existing models on the platform, and the documentation frames the addition as another selectable option rather than a default. That means enterprise teams can test it against other models already in the platform for quality, latency and cost before deciding where it fits best. The move also underscores how cloud AI marketplaces are evolving into neutral distribution channels for rival model makers. Instead of forcing customers into a single vendor’s ecosystem, platforms like Google’s are increasingly acting as aggregators, giving users access to models from multiple providers under one set of enterprise controls. For now, Grok 4.6’s presence on Google’s enterprise platform is likely to be watched closely by developers who need large context windows and by organizations already experimenting with agentic workflows. The combination of broad availability, configurable reasoning and enterprise distribution could make it a notable option in a crowded market for advanced AI models.

Nic Reeve·
What’s Really Known About OpenAI’s Rumored Next Model
AI & Tech

What’s Really Known About OpenAI’s Rumored Next Model

OpenAI’s rumored next model is generating headlines, but the most defensible version of the story is narrower than the leak chatter suggests: there is no official GPT-6 announcement , and much of what is circulating remains unverified. What appears to be real is that OpenAI is working on a new frontier system under a different name, while the specific claims about “GPT-6” remain speculation. Recent reporting from AI-focused outlets says the company’s next major model is being discussed internally under the codename Astra , not GPT-6. Those reports also say OpenAI has not published a model card, pricing, public API route, or release date, and has not confirmed whether the model will ship as GPT-6, as another GPT-5.x update, or under a different product line entirely. In other words, the name “GPT-6” is being used by the public and by leakers, but it is not an official label. The strongest verifiable signal comes from reporting that OpenAI has shown or discussed progress on a new model family while also slowing parts of development for safety review. One report says OpenAI paused reinforcement learning training for its latest-generation model for about two weeks to upgrade security, monitoring, and alignment systems. That report also says preliminary findings showed the model could identify and exploit a previously unknown zero-day vulnerability in the Artifactory software package proxy service. If accurate, that would explain why OpenAI may be taking a more cautious path before any public release. What should readers believe, then? First, believe that OpenAI is still actively working on a major next-step model. Second, believe that the company has not confirmed that the model is called GPT-6. Third, treat exact claims about parameter counts, context windows, release timing, or benchmark jumps as unverified unless OpenAI itself publishes them. The internet is full of confident numbers, but the available reporting does not support most of those specifics. The current rumor cycle has focused on a few recurring claims: a massive jump in training scale, a much larger context window, and a dramatic leap in reasoning or agentic behavior. Yet none of those details has been confirmed by OpenAI. Even articles that argue a new model is imminent pair that language with strong caveats that no official release date, product page, or spec sheet exists. That distinction matters. A credible leak can indicate the direction of a project, but it is not the same as a product launch. There is also a broader pattern here. OpenAI has recently become more visible in discussing frontier-safety issues, and that creates a natural gap between internal experimentation and public rollout. If the company believes a model could uncover serious security vulnerabilities or behave unpredictably in cyber-related tasks, then additional evaluation would be expected before deployment. That does not prove a delay, but it does make the safety-review explanation plausible. For readers trying to separate signal from noise, the best rule is simple: only the company can confirm the model’s name, timing, and capabilities . Until that happens, “GPT-6 leak” is better understood as a label for a bundle of rumors around OpenAI’s next frontier system, not as a verified product announcement. That means the most credible takeaway is also the least sensational one. OpenAI is almost certainly building something new, but the exact branding and release plan are still unclear. The more specific the claim — whether it is about the model’s size, performance, or launch date — the more skeptical readers should be unless it appears in official OpenAI material or in reporting with direct sourcing. In short, the safe position is not that GPT-6 is fake, but that GPT-6 is unconfirmed . The next OpenAI model may eventually earn that name, or it may arrive as Astra or under some other label. For now, the only thing that can be said with confidence is that the speculation has moved faster than the evidence.

Nic Reeve·
Anthropic Gives Claude Cowork Shared Memory with Chat for Persistent Context
AI & Tech

Anthropic Gives Claude Cowork Shared Memory with Chat for Persistent Context

Anthropic is rolling out a major upgrade to its AI assistant, giving Claude Cowork the ability to seamlessly reuse information it learns in regular chat. The company has merged the memory systems behind Claude’s chat interface and its Cowork desktop agent, so details you share in one surface can now automatically be used in the other. One Shared Memory Across Chat and Cowork Previously, Claude’s long‑term memory was largely confined to chat sessions and was synthesized periodically, meaning it could take up to a day before information carried over into new conversations. Cowork, which runs complex, multistep jobs on a user’s desktop or in the cloud, relied on its own background memory file and prompt stitching to simulate continuity. With the August 25 update, Anthropic has combined these mechanisms into a single, shared memory system that serves both chat and Cowork. Anthropic describes the change simply: the same memory now powers both Claude chat and Cowork. When users hand a task to Cowork—such as drafting reports, updating spreadsheets, or coordinating project documents—the context Claude has accumulated over months of chats is immediately available. Likewise, any new facts or preferences learned during Cowork runs are written back into the shared memory and become available in subsequent chat sessions. Real‑Time Memory, Not Just End‑of‑Chat Summaries Another important shift is how Claude updates memory. Instead of waiting to summarize an entire conversation once it ends, Claude now adds topics to memory in real time as users chat. This means that if a user mentions that a project deadline moved to September, that update can be reflected in memory almost immediately and show up in the very next interaction—whether in chat or Cowork—without requiring a manual “remember this” command. Anthropic’s support materials explain that when Cowork runs in the cloud, what Claude remembers from previous chats is automatically available, and what emerges during Cowork tasks feeds back into chat memory. Behind the scenes, each Cowork prompt is assembled from the user’s immediate request, their global instructions, and a relevant slice of the shared memory, allowing the AI to behave as if it has persistent awareness of roles, projects, and preferences. What Users Gain: Less Repetition, More Continuity The practical effect for users is that they no longer need to repeatedly brief Claude on who they are, what they are working on, or how they like to work every time they switch between chat and Cowork. Anthropic and independent commentators highlight several common scenarios: Persistent project context: Ongoing details such as quarterly goals, client names, and current project status can be retained across weeks or months and recalled in both chat and Cowork. Stable roles and preferences: If a user identifies themselves as an investment analyst, a teacher, or a particular type of creator, Claude can remember that role and tailor responses accordingly, even when individual chats are short or focused on different tasks. Cross‑device consistency: The shared memory applies across web, desktop, and mobile experiences, so moving from a browser chat to the Cowork desktop agent no longer breaks context. Tech industry observers note that this update positions Claude more directly as an AI “teammate” that can track medium‑ and long‑term workstreams instead of acting purely as a session‑bound chatbot. Transparency and User Control Over Memory The shared memory system arrives alongside a push for greater user control. Anthropic now surfaces everything Claude remembers in a dedicated Topics view within memory settings, where users can inspect, edit, or delete individual entries. Memory is stored as discrete, categorized entries rather than a single opaque summary, making it easier to remove outdated or inaccurate information. Users can also pause memory or reset it entirely if they no longer wish Claude to retain prior context. In addition, Anthropic provides guidance on importing and exporting memory, so the information Claude stores about a user is not locked in and can in principle be backed up or moved. Handling Sensitive Topics Anthropic has emphasized that the system is designed to minimize the capture of highly sensitive information by default. Topics such as health data, beliefs, and other potentially sensitive categories are excluded from memory unless users explicitly opt in via an “Include sensitive topics in memory” setting. For business customers, team or enterprise administrators can centrally control whether memory is enabled at all, and may choose more restrictive policies depending on corporate governance requirements. External reporting indicates that memory generation is turned on by default for free, Pro, and Max plans, while Cowork itself is not available on free accounts. For organizations that want to keep different workstreams separated, Anthropic has indicated that the only way to maintain fully separate memories for chat and Cowork is to use different accounts, since the new system treats them as a single unified space. Availability and Limitations The new shared memory capability began rolling out on August 25, 2026, across Claude’s web, desktop, and mobile experiences, as well as Cowork running in the cloud. Earlier in the year, memory support was limited to chat surfaces, and some third‑party analyses noted that Cowork lacked access to that long‑term context. Anthropic’s latest release notes and help center now explicitly state that memory works across both chat and Cowork when the latter runs in the cloud environment. There are still technical constraints. Cowork’s use of memory depends on cloud execution rather than purely local processing, and incognito or memory‑disabled sessions remain stateless by design. As with other AI systems, Anthropic cautions that Claude’s memory is selective: it prioritizes high‑level preferences and recurring topics rather than storing every detail of every conversation. A Step Toward More Personalized AI Workflows By unifying memory between Claude chat and Cowork, Anthropic is betting that users will value a more personalized and continuous AI experience, particularly for complex, ongoing work. The update reduces friction for individuals juggling multiple projects and gives enterprises a clearer path to building AI‑augmented workflows that persist over time. At the same time, the company is attempting to balance convenience with privacy and security by giving users fine‑grained controls and limiting sensitive data retention by default.

Nic Reeve·