Apple and X SpaceX Settle Antitrust, AI Chips Propel Broadcom Growth, M3E Canvas Converts Visuals to AI Code Prompts, AI Giants Propose Peer Review Safety, and more...

0:00Good morning, it's September 16th and this is your daily brief in AI. Here's everything you need to know. A settlement has been reached in the antitrust dispute involving X-Corp-slash-SpaceXAI and Apple, while a related case against OpenAI continues. The dismissal of the Apple-slash-X.AI claims leaves settlement terms undisclosed and the filing provides no reasons for the dismissal. OpenAI also faces a separate lawsuit from Apple overtrade secret issues tied to unreleased products, adding pressure amid broader regulatory scrutiny. The article notes much of the content is syndicated, with elements like the headline and image possibly revised. Promotional affiliate disclosures and links appear, but they are ancillary to the core legal news. The reporting situates tech elite dynamics and political considerations, hinting at implications for musk-backed ventures in the ecosystem. Disclosures is the attribution for the piece, with the date tied to a mid-September

1:002026 filing and reporting window. Mike Scarcella of Reuters is cited, highlighting a September filing from Washington and ongoing legal action. Inputs from multiple agencies are acknowledged, and settlement terms, if any, have not been publicly disclosed. Analysts view the resolution in the context of heightened antitrust scrutiny over AI distribution deals and exclusive integrations that could affect competition. European regulators have not designated virtual assistance as a core platform service under the DMA, with ongoing assessments into AI service classifications set for review. XCorponX.AI have been reorganized under SpaceX, aligning musk's AI ventures within the broader corporate structure. A sweeping AI hardware wave is reshaping the semiconductor landscape, driven by surging AI revenue that now dominates the mix. AI semiconductor revenue jumped 211% year over year to $116.7 billion, making up 56% of

2:03total revenue, and up from 49% last quarter, as the AI hardware boom continues to lead the revenue mix. The company raised its fiscal 2026 AI revenue outlook to $58 billion, about 186% above the prior year, underscoring confidence in AI-driven growth beyond current guidance. Semiconductor solutions hit a record $20.8 billion in revenue, up 127% year over year, and representing 70% of company sales, supported by a 61% segment margin driven by XPS and AI hardware. Deployment roadmaps with anthropic and open AI point to gigawatt-scale AI infrastructure through 2028, underpinning demand for ironwood, TPUV8i, Chalepeno, and future accelerators. The company expanded its partnership with Google to co-develop and supply next-generation TPU and AI networking, with multi-billions in annual TPU deployments planned.

3:04For the quarter, revenue reached $29.6 billion, up 86% year over year, with operating income of $20.1 billion, up 92%, and free cash flow of $13.7 billion. Operating margin rose to 67.9% of revenue, and non-GAP earnings per share increased 96% year over year to $3.32, reflecting operating leverage. AI hardware winds include ironwood TPUV7 with anthropic and Google, Google TPUV8i, and open AI's Hallipenio accelerator rollout. The AI mix is led by custom accelerators, notably XPS, which accounted for 73% of AI revenue with six large XPU customers, contributing $116.7 billion, more than triple prior levels. The long-term trajectory targets AI semiconductor revenue near $230 billion by 2028, with potential

4:07$28 earnings per share above $30. A new browser-based UI canvas is cutting design to code friction by turning visual designs into precise prompts for AI coding agents, speeding front-end prototyping and development. The tool, M3E Canvas, is open source and focuses on material three expressive interfaces, adding you visually sketch mobile and desktop layouts and export structured prompts for AI tools. It runs entirely in the browser with no back end, built on next.js16 and React 19, and supports arranging components like app bars, fabs, chips, cards, sliders, and navigation rails, plus flow prototyping to validate user flows. Quickstart guidance is provided to run locally by cloning the GitHub repository, installing dependencies, and starting the DevServer with links to the live app and repository for exploration and contribution. The Canvas includes a rich material expressive catalog with magnetic snapping, shape morphing

5:10loading indicators, and responsive layouts that adapt between mobile and desktop widths. Interactive navigation is supported through tap and swipe, with configurable transitions and direct navigation targets to preview user flows before code generation. A Corsi is prioritized with 100% client-side operation and no telemetry. Designs are stored in local storage, and optional local AI helpers can use the user's own API keys for Claude, OpenAI, or Gemini to write in-browser component behavior notes. A core feature is structured AI prompt generation. A single action compiles the visual scene into a deterministic prompt that specifies component hierarchies, layout margins, target text acts, Android compose or modern web, and multilingual prompts in English, Japanese, Chinese, and Korean. This combination aims to streamline how designers and developers move from visuals to code-ready prompts without leaving the browser.

6:15A new mutual verification idea aims to have leading AI labs test each other's frontier models for safety using a shared test harness, a peer review style check before public release. The proposal envisions participants like XAI, OpenAI, Anthropic, Google, Meta, and Top Chinese firms collaborating to run standardized tests. Implementing this faces hurdles such as intellectual property concerns and regulatory differences between the United States and China, with proposed solutions including encrypted testing environments and third-party oversight. The timing comes as debates continue over balancing rapid development with safety, with voices calling for slower, more cautious progress. Peer testing is unusual in this fast-moving sector, but supporters say it could influence how companies weigh safety against commercial and national security interests. CNBC notes the plan isn't yet agreed to by rivals, and some leaders have floated alternatives like third-party embedded evaluators.

7:17The core idea centers on an industry self-discipline framework, opening APIs to competitors and conducting independent security testing with shared logs and open source tools to deter information leakage. If widely adopted, the approach could standardize safety protocols and shape global AI governance, potentially rewarding early adopters with strong regulatory compliance and public trust. A key mechanism would have competitors flag safety concerns and raise alarms about dangerous behavior rather than each lab auditing its own model. Analysts see potential for better detection of risky behavior before release, while acknowledging tensions over intellectual property and security. Broughter tensions between the United States and China over slowing AI development could influence whether a mutual verification framework gains traction, and some leaders argue external scrutiny could substantially increase the chances of spotting problems, even if the plan isn't perfect.

8:18AI scaling goes beyond technology alone. It requires full business transformation with governance, trusted data, and human oversight to succeed. Infosys TOPAS is designed to move generative AI into practical business applications, boosting productivity and operational efficiency across the enterprise. For piloting AI, leaders should define the problem, data sources, risks, and success criteria to test a focused use case before broader rollout. A real-world example from Tennis Australia shows how match feel enables tactile accessibility and AI-driven insights at the Australian Open when aligned with user needs and human expertise. AI initiatives should be treated as ongoing programs with clear business ownership governed by responsible AI practices rather than isolated experiments. Enterprise AI success depends on people and operating models as much as technology, with research noting only a minority of use cases meeting most business objectives. Post-apployment AI systems require continuous monitoring for accuracy, security, bias,

9:24and unintended consequences, with risks understood and bounded. Learning and AI literacy across the workforce is essential, so employees understand value, usage, and when to rely on human judgment, not just IT-led deployment. AI data must be accurate, secure, and appropriate, with strong governance and fairness measures enabling safer, more efficient deployment without stifling innovation. Infosys' responsible AI framework and ISO4-201 certification illustrate governance and management standards for Enterprise AI. The shift from pilot to scale hinges on measurable business outcomes, reliable data and integration, and clear ownership of results across leadership and teams. AI should fit into existing workflows and be part of a broader business transformation program rather than standalone tools. This has been your daily brief in AI. To read more about these stories, follow the links in the episode bio.

10:26You can also subscribe to these updates via email at www.brief.News. For more daily podcasts about the topics you love, visit brief.News-forward slash podcasts. Tune in tomorrow. We'll be back with everything you need to know.