Was the Microsoft 365 outage a hack? Microsoft says no, it was its own servers, and its Copilot boss “regrets” it

windowslatest.com Nov 13, 2026

Was the Microsoft 365 outage a hack? Microsoft says no, it was its own servers, and its Copilot boss “regrets” it

Curated by Shubham · TechCapsules · Nov 13, 2026

AI-summarised brief · reviewed before publication

Microsoft confirmed that the worldwide Microsoft 365 outage, which began at 8:30 a.m. PT on November 12, was caused by overloaded internal servers rather than a cyberattack. Executive Vice President Charles Lamanna, head of Copilot, Agents and Platform, said the company “deeply regrets” the disruption that left millions unable to access Outlook, Teams, SharePoint and Copilot. Microsoft’s health center timeline shows the issue escalated as servers became unavailable, prompting traffic rerouting and a configuration change that began stabilizing services by mid‑afternoon, though full capacity remained unrecovered into the early hours of November 13. The outage, the worst in the firm’s history, continues to affect users despite patches and health checks, prompting calls for SLA‑based compensation.

💡 Why It Matters

· The incident exposes the fragility of cloud‑centric work ecosystems when a single provider’s infrastructure falters.

Read full article on windowslatest.com

Outlook, Teams, Copilot, and more are inaccessible for millions

Microsoft’s cloud platform suffered a worldwide outage that began around 1:00 p.m. Eastern on Thursday, November 12, 2026, and persisted for more than 17 hours, disabling Outlook, Teams, Copilot and the entire Exchange service for millions of users. The disruption spread across multiple Microsoft 365 applications, prompting hourly status updates that initially reported slow repair progress and later noted “early signs of stabilization.” By 6:00 a.m. Eastern the company said recovery actions in the Asia‑Pacific region had restored some traffic and improved mailbox access, though intermittent problems remained. Microsoft emphasized no evidence of malicious activity and pledged continued mitigation. The incident follows an October change to Exchange Web Services that had previously affected Apple Mail. Enterprise customers reported lost productivity while small businesses struggled to meet deadlines. Many firms faced delayed communications and halted project workflows today.

💡 The outage exposed the fragility of corporate reliance on a single cloud provider, forcing organizations to reconsider continuity plans while highlighting the cascading risk of a unified service architecture.

9to5mac.com · Nov 11

Apple confirms ongoing outage affecting iCloud and more

Apple confirmed that a widespread outage began at 5:49 p.m. ET/2:49 p.m. PT, disrupting iCloud, Find My, and related services. Users reported being unable to send or receive iCloud email, sign into their Apple accounts, and in some cases were logged out entirely. The company acknowledged the issue on its System Status page and stated that Find My might be “slow or unavailable.” Apple announced that all outages have now been resolved.

💡 The interruption exposed the fragility of Apple’s cloud infrastructure, a backbone for millions of daily transactions and personal data. It underscored the company’s reliance on uninterrupted service for user trust and operational continuity.

thehackernews.com · Nov 9

Hackers Exploit Qubole’s Trusted AWS Access to Breach Customer Accounts and Steal…

Between October 14 and 26, attackers compromised AWS accounts belonging to multiple Qubole customers, stealing stored secrets and data, according to cloud‑security firm Wiz. The breach exploited the trust relationship that Qubole’s platform establishes with customers via an IAM role that permits Qubole’s AWS account (805246085872) to assume the role using an External ID. The attackers obtained credentials for that Qubole account—method undisclosed—and successfully issued AssumeRole requests from outside AWS, gaining access to unused but still trusted roles. Once inside, they harvested environment variables, Secrets Manager entries, SSM Parameter Store values, and copied data from any S3 buckets the roles could reach, sometimes downloading entire buckets. Wiz found no link to known state‑backed campaigns and noted that both Qubole and its owner Idera have not issued statements.

💡 The incident shows how lingering trust relationships and non‑secret External IDs can let attackers pivot across multiple clients, turning a single compromised service account into a multi‑tenant data breach vector.

9to5google.com · Nov 5

Madden NFL 27 and NBA The Run are now free on Amazon…

Amazon Luna has added EA’s Madden NFL 27 and NBA The Run to its free library, available today for users with an active Amazon Prime membership. The sports titles, released in August, are now part of Luna’s Prime perks, allowing players to stream the games without additional cost. Amazon also announced the release of the “MADDEN” movie on Prime Video, scheduled for November 18, tying the gaming and streaming offerings together. The move expands Luna’s sports catalog amid the NFL season and NBA playoffs.

💡 By bundling popular sports games with Prime, Amazon strengthens its competitive edge in cloud gaming, offering immediate, cost‑free access that could sway casual players toward its platform.

thequantuminsider.com · Nov 5

Quantinuum Raises Revenue Outlook After Landing Its Largest Cloud Agreement

Quantinuum announced third‑quarter revenue of $9.1 million, a 233 % increase from the same period a year earlier, and lifted its full‑year revenue guidance to $31‑34 million. The company also secured its largest cloud contract, a deal with a leading financial institution that guarantees access to Quantinuum’s planned 2029 Apollo quantum computer and its Nexus software platform. Nexus usage now spans 215 organizations, while new agreements were signed with a pharmaceutical firm, GE Vernova, and Abu Dhabi Maritime Academy, expanding applications in drug discovery, superconducting materials and maritime logistics. Quantinuum reported error‑correction fidelity of 99.9995 % on its Helios processor and continued deployments in Singapore, Japan and U.S. defense projects. Despite the growth, the quarterly net loss widened to $118 million, though cash reserves stood at $1.94 billion.

💡 The multiyear cloud deal secures early revenue from Quantinuum’s future hardware, giving the company a cash‑flow advantage that few quantum rivals can match.

techcrunch.com · Nov 5

Serve Robotics launches new business to become the AWS for robots

Serve Robotics, the maker of autonomous sidewalk delivery robots, announced a new business unit called Physical AI Services aimed at supplying connectivity and data tools to other robot developers. The move follows its 2021 spin‑out from Uber’s Postmates acquisition and its 2024 IPO. Early customers include Bedrock Robotics, which builds self‑driving systems for construction equipment, and a partnership with T‑Mobile to bring edge AI computing closer to robots. Physical AI Services will launch three products: Serve Link, a high‑performance link that lets robots tap cloud‑based AI models; a data curator that surfaces edge‑case footage for model training; and a foundation‑model data catalog derived from Serve’s own delivery fleet. The company will continue operating its delivery robots and hospital units while expanding the platform to farms, warehouses and other uncontrolled environments.

💡 By packaging its in‑house robotics stack as a cloud‑like service, Serve Robotics could become the de‑facto infrastructure layer for a fragmented market of niche robot makers.