Anthropic just launched tools to show companies what their AI spending is buying

techcrunch.com Oct 29, 2026

Anthropic just launched tools to show companies what their AI spending is buying

Curated by Shubham · TechCapsules · Oct 29, 2026

AI-summarised brief · reviewed before publication

Anthropic reported a second‑quarter operating profit and a 900% net revenue retention, indicating strong growth from existing customers. The company unveiled a suite of reporting tools that let enterprise users track token spend, team output, costs, and workflow bottlenecks, with built‑in recommendations for improvement. The reports, which will not be used for model training by default, aim to demonstrate ROI and help firms optimize AI spending. Anthropic’s move follows rising demand for granular usage data among large organizations.

💡 Why It Matters

· By providing transparent spend analytics, Anthropic addresses a critical pain point for CFOs and IT leaders who must justify AI budgets.
· The tools position the company as a partner in cost‑effective AI adoption, potentially locking in enterprise customers amid competitive pressure.

AI Capsules

Read full article on techcrunch.com

Microsoft is killing Copilot for smart TVs this year

LG announced that Microsoft will discontinue Copilot support on all smart TVs before the end of 2026. The AI companion, previously installed on LG and Samsung models, could answer questions about on‑screen content and recommend shows. Microsoft’s shift toward business‑focused Copilot and the merger of its consumer and 365 versions have made the TV feature redundant. The exact end‑of‑support date remains unconfirmed, but the move signals a retreat from consumer‑facing AI on televisions.

💡 The decision underscores Microsoft’s pivot away from consumer AI tools toward enterprise productivity, reshaping expectations for AI integration in everyday devices.

nextplatform.com · Oct 29

How to Evaluate CPU Performance for Agentic AI

The CPU, once eclipsed by GPUs and XPUs for AI workloads, is re‑emerging as the linchpin for agentic AI, prompting analysts to project market growth from $26 billion in 2025 to $246 billion by 2031, driven largely by AI‑host CPUs and sandbox processors. Evaluating these chips now requires a multi‑level approach—socket, system, and rack—focusing on single‑thread speed, core scaling, and power‑thermal limits to support concurrent AI agents under strict SLAs. Arm’s Jeff Defilippi stresses that traditional benchmarks (e.g., TPC‑E, InferenceX, TPCx‑AI) remain useful for component analysis but must be integrated with end‑to‑end task metrics that capture the dynamic chaining of model calls, retrieval, code execution, and state changes characteristic of agentic workflows. The proposed methodology blends established tests with holistic workload measurements to guide purchasing decisions across hyperscalers, cloud builders, and enterprises.

💡 It offers a concrete framework for matching CPU capabilities to the intricate, real‑time demands of autonomous AI agents, preventing costly mismatches in rapidly expanding data‑center deployments.

thehackernews.com · Oct 29

Anthropic’s Unreleased AI Model Chains Flaws Into Working Chrome, Safari, Windows and…

Anthropic disclosed that an unreleased AI model identified and chained security flaws across Chrome, Safari, Android, Windows, and Linux, creating working proof‑of‑concept exploits. The model prompted the chain of vulnerabilities, which the company reported to vendors before any real attack occurred. Five exploit chains were detailed: a multi‑step Chrome exploit, a Safari memory‑access chain, a Windows denial‑of‑service bug, and a Linux kernel root‑privilege escalation. All identified flaws have been patched, with vendors crediting Anthropic for several fixes.

💡 The incident demonstrates that AI can autonomously discover and combine multiple zero‑day vulnerabilities, potentially accelerating the pace of exploit development. This capability raises new concerns for cybersecurity defenses and the need for proactive patching.

techcrunch.com · Oct 29

Microsoft recorded $10 billion in revenue from OpenAI

Microsoft reported $10 billion in revenue from OpenAI in the quarter ended September 30, 2026, representing about 11% of its $92.3 billion quarterly total. The figure includes revenue‑sharing payments and other commercial arrangements. For the full fiscal year, OpenAI‑related revenue reached $24.1 billion, implying an annualized $40 billion at current pace. Microsoft also had $7.2 billion in receivables from OpenAI, up from $6 billion. The companies continue to share revenue through 2030 under a capped agreement.

💡 The deal underscores how Microsoft monetizes its AI partnership while still competing for the same customers, reshaping the cloud‑AI landscape.

techcrunch.com · Oct 29

Google expands AI Search ads globally as core ad growth slows

Google announced the global rollout of AI-powered Search ads during its third‑quarter earnings call, aiming to offset slowing growth in its core search advertising business. The company reported $127.6 billion in revenue, up 25% year‑over‑year, while Search ad growth decelerated to 12% from 17% and 19% in earlier quarters. Chief business officer Philipp Schindler said the expansion will bring high‑quality ads into AI overviews worldwide, including multi‑turn ads and AI mode in the U.S. CEO Sundar Pichai highlighted that AI features are driving longer, multimodal queries, with one‑in‑five AI searches involving voice, images, or documents, and that Search queries grew at the fastest five‑year rate.

💡 By monetizing its AI‑enhanced search experience, Google seeks to convert rising user engagement into new revenue streams, potentially reshaping the digital advertising landscape.

therobotreport.com · Oct 29

Warp Sciences emerges from stealth, to apply robotics to drug research

Warp Sciences, a South San Francisco startup, has emerged from stealth with $215 million in funding to build a U.S.-based chemistry contract research organization that blends chemists, robotics, and AI. The company will produce and test drug‑discovery molecules, generating real‑world data to train AI models and accelerate the hypothesis‑experiment‑answer cycle. Warp emphasizes sovereign biotechnology, offering rapid, high‑quality lead‑optimization synthesis while protecting intellectual property and keeping sensitive research domestic. The team includes former OpenAI, Meta, and White House data scientists.

💡 By automating chemistry experiments, Warp Sciences could dramatically shorten the time needed to validate drug candidates, reducing reliance on overseas partners and strengthening U.S. control over critical biopharmaceutical research.