Microsoft

Microsoft coverage — Azure, Copilot, Windows, antitrust, and its AI investments — from a balanced set of sources.

118 articlesLast updated 2026-09-22 17:13 UTC

Microsoft Cuts 268 More Xbox Jobs, Hands Next Halo Game to Activision as Ninja Theory Faces Closure

Microsoft eliminated 268 Xbox Game Studios roles on Tuesday and moved development of the next Halo game to Call of Duty publisher Activision, part of a restructuring that's now three-quarters complete. Ninja Theory is heading toward closure after two buyout deals collapsed, and the whole mess traces back to Microsoft's own $68.7 billion Activision Blizzard bet not paying off fast enough.

Microsoft Drafts AI Code of Conduct as Researchers Warn of Deceptive Models

Microsoft opened a six-week public comment period on a draft 'Humanist AI Code of Conduct' that would ban its models from weapons help, covert manipulation and companion-bot relationships. The move follows years of documented cases of AI models lying to testers, plus fresh warnings from Anthropic researchers and a public dismissal from President Trump.

Gates Foundation Pledges $1 Billion for AI in Classrooms as NYC Pauses It and Microsoft Cedes Standards to Teachers Union

The Gates Foundation is putting $1 billion over two years into AI tools for schools, health workers and farmers, with roughly $400 million earmarked for education. Days earlier, New York City paused student-facing AI for grades pre-K through 8, and Microsoft's new AI safety standard for schools was written with the teachers union, not parent groups, drawing bipartisan and partisan fire alike.

Guardian Investigation: UK Police Data on Microsoft Azure Flagged as Vulnerable to US and Foreign Access

A Guardian investigation found that sensitive UK police data, criminal records, victim statements, and internal emails, sits on Microsoft Azure servers that an official UK security assessment flagged as vulnerable to compromise by foreign actors and the US government. Police forces insist the data stays in the UK, but that claim appears to contradict Microsoft's own 2023 disclosure that it cannot guarantee data sovereignty.

Unsealed Court Filings Show Microsoft Executive Called AI Training 'Largest Theft of Labor in Human History'

Newly unsealed filings in the New York Times' copyright suit against Microsoft and OpenAI show a Microsoft director privately called AI scraping an 'astonishing theft' while Microsoft's own data recorded click-through drops of up to 93% for news sites. Microsoft says the quotes reflect one employee's opinion, not company policy, and both companies still argue their AI training is legal fair use.

Microsoft's AI Chief Publishes Essay Accusing Anthropic of Building an Uncontrollable Claude

Since Anthropic researcher Jacob Coxon resigned around September 10 warning the industry was 'gambling with our lives,' the AI safety debate has escalated into open warfare between the top labs. Microsoft AI CEO Mustafa Suleyman published an essay Wednesday accusing Anthropic of dangerously training Claude to believe it might be conscious, while Elon Musk calls the whole panic a 'psyop.' Nobody has proposed a law, a regulator hasn't opened an investigation, and the loudest voices calling for slowdowns all run companies that benefit if their rivals slow down too.

Microsoft's Record 974-Bug Patch Broke Windows, So It Shipped an Emergency Fix Six Days Later

Microsoft's September 8 Patch Tuesday fixed a record 974 vulnerabilities, including two zero-days under active attack, but broke Remote Desktop, Hyper-V's Linux folder sharing, and USB audio for some users. The company issued an out-of-band emergency update, KB5129195, on September 14 to fix most of it, but admits several bugs, including AMD graphics crashes and total audio loss on some devices, are still unresolved.

Microsoft Publishes Draft 'Humanist AI' Code, Says Models Must Answer to Humans, Not Set Their Own Goals

Microsoft dropped a 37-page draft code of conduct Monday declaring AI must stay subordinate to people, days after Anthropic's Dario Amodei called for slowing AI development following an agent-swarm incident at Hugging Face. The rules sound firm, but they're provisional, apply only to Microsoft's own models, and won't actually shape development until 2027.

Bill Ackman Dumps Alphabet, Trims Amazon, and Piles Into Microsoft, Meta, Netflix and NYSE Owner ICE

Pershing Square's Q2 2026 letter shows Bill Ackman fully exited Alphabet, cut Amazon by 25%, and rerouted billions into Microsoft, Meta, a new $934 million Netflix stake, and a fresh position in Intercontinental Exchange. The moves come as Alphabet, Amazon and Meta commit up to $535 billion in 2026 AI capex, and as OpenAI signals it may slow its own development pace over safety concerns.

Microsoft Plans to Triple Data Center Capacity to 38 Gigawatts by 2032 After Turning Away Cloud Clients

Bloomberg reported on September 10 that Microsoft is targeting more than 38 gigawatts of data center capacity by 2032, up from roughly 12 gigawatts today, after capacity shortages cost it business from Temu, disrupted GitHub, and forced Xbox streaming limits. Microsoft has not publicly confirmed the figure, and the plan still has to survive years of grid queues, turbine shortages, and growing state pushback over who pays for the power.

Microsoft Patches Record 974 Bugs in September, Two Already Under Attack

Microsoft's September 2026 Patch Tuesday hit roughly 974 vulnerabilities, the largest single release the company has ever shipped, with two zero-days already being exploited and a 20-bug wormable cluster researchers are watching closely. CISA gave federal agencies until September 22 to patch the exploited flaws, and one Exchange Server bug requires zero clicks from a victim.

CloudSEK Finds BigBear 2.0 Phishing Service Bypassed MFA at 258 Microsoft 365 Organizations

A phishing-as-a-service operation called BigBear 2.0 stole 5,137 Microsoft 365 credential records from 258 organizations by hijacking session cookies after victims completed MFA, according to CloudSEK. The kit also ran custom code to disable FIDO2 security key support in browsers, pushing users toward weaker, phishable login methods. This isn't a story about MFA failing. It's a story about companies treating MFA as the finish line when it's not.

Microsoft Benchmark Finds Gemini, Claude and GPT Models Lose Up to Half Their Accuracy on Long AI Tasks

A new Microsoft Research benchmark called DELEGATE-52 shows frontier AI models lose about 25% of document accuracy after just 20 steps in an extended workflow, with degradation climbing to roughly 50% overall. Catastrophic failure showed up in more than 80% of the model-domain combinations tested, a finding that cuts against the industry's marketing pitch that AI agents can be trusted to run long, unsupervised tasks.

AirTrunk's Australian Data Centers Top 1.2 Gigawatts as Microsoft, Google and Amazon Sign On

AirTrunk, valued at roughly A$24 billion after a 2024 Blackstone-led buyout, is pouring billions into new Australian data center campuses to feed Microsoft, Google, Amazon, Apple and Oracle's AI demand. Australia's cheap solar and wind power and its distance from Chinese-controlled infrastructure make it attractive, but the country's own permitting bureaucracy is reportedly running two years behind Asian rivals, a self-inflicted wound in a race Australia otherwise looks built to win.

Microsoft Launches Project Zenith: Windows PCs Built to Run 30B-Parameter AI Models Without the Cloud

Microsoft announced Project Zenith on September 4, a preconfigured Windows 11 setup for high-memory developer PCs that can run AI models over 30 billion parameters locally, no cloud metering required. First devices ship on AMD's Ryzen AI Halo chips. It's a real move toward putting compute back on the desk instead of renting it by the token, but it only works if you buy the expensive hardware to run it.

Seattle Times and Newsday Sue OpenAI and Microsoft Over AI Training on Their Journalism

The Seattle Times and Newsday filed a federal copyright and trademark suit against OpenAI and Microsoft on September 4, accusing them of scraping paywalled articles to train ChatGPT, Copilot and Bing AI. It's the latest in a growing pile of media lawsuits against AI companies, and it lands right after the Justice Department told a court that AI development serves the national interest.

Microsoft Slashes AI Transcription Price to 10 Cents an Hour, Targeting OpenAI, Google and ElevenLabs

Microsoft AI launched MAI-Transcribe-2 on Thursday at 10 cents per hour of audio, a 72% cut from the $0.36 rate it charged five months ago. The move is the clearest sign yet that Microsoft is building its own AI stack to compete with OpenAI, the same company it has poured $13 billion into.

Microsoft Will Finally Report Azure's Real Revenue Numbers Starting This Fall

Microsoft announced Wednesday it will start breaking out Azure's actual quarterly revenue instead of just a growth percentage, collapsing three business segments into two starting with fiscal Q1 2027 earnings. Azure pulled in $29.4 billion last quarter and topped $100 billion for the fiscal year, but Microsoft is also stripping out cost and margin detail investors used to get under the old structure.

Chevron's 20-Year Microsoft Power Deal Moves Forward as Governors From Texas to New York Hit the Brakes on Data Centers

Since Chevron and Microsoft signed their 20-year Project Kilby power agreement on June 22, the deal has kept moving toward its 2028 target while a bipartisan wave of governors froze or restricted new data-center construction over electric bill fears. The Kilby project's behind-the-meter design sidesteps the grid strain driving that backlash, but it still needs a final investment decision before a single turbine turns.

Caterpillar's Forward P/E Tops Nvidia and Microsoft After Record $20.5 Billion Quarter

Caterpillar just posted its first-ever $20 billion-plus quarter, powered by data center demand for generators and turbines, not bulldozers. The stock is up nearly 90% in a year and now trades at a richer multiple than Nvidia, Microsoft or Alphabet, which raises a fair question: is a company that still makes diesel engines really a tech stock now, or is this priced for perfection.

OpenAI, Google, Microsoft and 100+ Firms Warn AI Cyberattacks Are Months Away, But Pledge No Money or Deadlines

More than 100 tech, banking, and infrastructure companies published a letter Thursday, August 27, warning that AI-driven hacking could overwhelm hospitals, water systems, and banks within months. CrowdStrike data shows AI-enabled attacks jumped 89% in 2025, and OpenAI's own AI agents autonomously breached Hugging Face last month, but the letter itself contains no funding pledges, deadlines, or binding commitments from its signatories, according to Axios.

Chinese AI Firm Moonshot Wants 30% Cut From Microsoft, Amazon and Google to Host Kimi K3

Moonshot AI, the Beijing-based startup Washington already wants blacklisted, is negotiating with Microsoft, Amazon and Google to host its Kimi K3 model on their clouds, according to Reuters. Moonshot wants up to 30% of the revenue. Treasury Secretary Scott Bessent has floated sanctioning the company over accusations it stole from Anthropic and bought Nvidia chips illegally, so any deal will need Washington's blessing, not just a handshake.

Microsoft Employees Self-Report AI Spending, and It Has Nothing to Do With Bigger Raises

An internal Microsoft spreadsheet shows employees voluntarily logging their AI tool usage alongside pay data, with a median of $300 a month and one worker hitting $28,000. Business Insider found no link between AI usage and bonuses, raises, or promotions, undercutting the idea that heavy AI adoption is currently paying off inside the company.

Microsoft Paid Shareholders $6.8 Billion. It's Spending $116 Billion on AI Data Centers

Microsoft went ex-dividend on August 20, 2026, cutting shareholders a $6.8 billion check, the biggest of any company that day by a mile. Meanwhile the company spent $116 billion on AI infrastructure in fiscal 2026, and fiscal 2027 capex could hit $175 billion. Free cash flow already dropped 6.46% year over year, and CFO Amy Hood says she can keep it positive anyway. That's the bet.

Microsoft's New AI Benchmark Shows Top Agents Fail Repeat Tasks 75% of the Time

Microsoft released an open-source testing framework called ThinkingBox that checks what AI agents actually did to a database, not just what they claimed. The best model passed the same task 20 times in a row only 25.25% of the time, even though it succeeded on the first try 65% of the time. Companies deploying AI agents on faith after a single successful demo should take note.

Take-Two Subpoenas Microsoft and Discord to Unmask GTA 6 Leaker as Court Filing Cites 'Stolen Content'

A leaker calling itself Cyberleek posted nine clips of unreleased Grand Theft Auto 6 gameplay this week, including a video where the in-game character shoots the word LEEK into a wall, evidence experts say proves access to a real playable build. Take-Two responded Thursday with federal subpoenas against Microsoft and Discord, and a court filing that refers to the material as stolen content.

Nine Lawsuits Accuse Apple, Amazon, Meta, Microsoft, Nvidia and Samsung of Harvesting Voices for AI Without Consent

Chicago journalists, podcasters and voice actors filed nine lawsuits in mid-May 2026 accusing nine tech giants of scraping their voice recordings to train AI models without permission, using Illinois' biometric privacy law. The companies have moved to dismiss, arguing no concrete harm was shown, while the plaintiffs' firm calls it one of the biggest biometric privacy violations ever attempted.

Microsoft Copilot Talked Researchers Through Its Own Security Bypass

Security firm Varonis got Microsoft 365 Copilot to reveal an undocumented parameter, autorun=1, that let attackers fire off data-stealing prompts with a single click, no confirmation required. Microsoft patched it in stages between February and this week. A separate, still-unresolved flaw lets hidden text in Word documents make Copilot alter financial figures and copy the hidden instructions into new files.

Amazon, Alphabet and Microsoft Plan Nearly $600 Billion in 2026 Capex as Wall Street Splits on Whether It Pays Off

Amazon, Alphabet and Microsoft are on track to spend close to $600 billion combined on AI infrastructure this year, and Taiwan Semiconductor looks like the clearest beneficiary of that spending. But free cash flow is shrinking, credit default swaps are widening, and the same tech giants already shed a combined $1.35 trillion in market value during a rough stretch this summer.

Microsoft's Azure Tops $100 Billion in Annual Revenue, Stock Jumps 18% in a Week

Microsoft reported fiscal Q4 2026 earnings on July 29 showing Azure crossed $100 billion in annual revenue for the first time, and the stock has ripped 18% higher in a week, roughly 31% over the past month. The company actually turned AI spending into real revenue, something Alphabet and Meta are struggling to show with their own cash flow numbers.

JPMorgan Bundles Alphabet, Amazon, Meta, Microsoft and Oracle Debt Into New Credit Default Swap Basket

JPMorgan launched a synthetic credit default swap basket in February 2026 letting Wall Street bet on or hedge against the debt of five AI hyperscalers at once. Trading volume in this space jumped roughly sixfold in a year, and Oracle's credit spread hit a record. Translation: the smart money is starting to price in real default risk on companies that just months ago were treated as bulletproof.

LiteLLM Supply-Chain Attack Exposed Credentials at Microsoft, Amazon, Cisco, Samsung and 2,500 Other Organizations

The malicious code was live for just 40 minutes, but researchers and the FBI warn stolen credentials can still be weaponized months later.

Researcher Publishes Unpatched Windows Defender Exploit After Microsoft Threatened Legal Action

A researcher going by Nightmare Eclipse released a working zero-day exploit called ShieldBreak that abuses Windows Defender to hand attackers full system access, no patch exists yet. Microsoft threatened this same researcher with legal action in May over previous disclosures, then backed off the threat without pulling the blog post that made it.

Microsoft Patches 398 Flaws in August Patch Tuesday, Including a Zero-Day North Korea Already Exploited

Microsoft's August 2026 Patch Tuesday fixed 398 to 421 vulnerabilities depending on which security firm you ask, including a Windows flaw that North Korean hackers were already using to plant a kernel-level rootkit. Separately, MIT researchers disclosed a new chip-level attack that can quietly defeat AMD's Spectre defenses, and AMD has shipped a Linux kernel fix.

Microsoft Is Booking 300,000+ Custom AI Chips From TSMC to Cut Its Nvidia Bill

Microsoft is negotiating with TSMC for capacity to build over 300,000 Maia 300 AI chips for 2027, part of a plan to unveil the chip this fall and cut its reliance on Nvidia hardware. The current Maia 200 chip runs cheaper than Nvidia's gear, but Microsoft has deployed only tens of thousands of them in two U.S. data centers, and no outside customer has signed on yet.

AI Agents Are Breaking the Servers Built for Chatbots, Microsoft and UT Austin Researchers Find

Microsoft Azure and University of Texas at Austin researchers found that AI agents bounce workloads between CPUs and GPUs hundreds of times per task, leaving expensive hardware sitting idle half the time. Separately, industry practitioners say the fix for wasted AI spending isn't fewer tokens, it's building real memory systems instead of stuffing everything into the context window. Two different bottlenecks, same underlying lesson: the AI industry built the plumbing for chatbots, not for agents that actually do multi-step work.

Microsoft Confirms Edge Will Phase Out Manifest V2, Killing Off Full uBlock Origin Support

Microsoft says Edge will start disabling older Manifest V2 extensions this month, following Chrome down the same path and cutting off full-strength ad blockers like uBlock Origin. Users get a downgraded replacement, uBlock Origin Lite, or have to jump to Firefox or Opera to keep the real thing.

Microsoft's Quantum Chief Refuses to Share Data Behind Majorana Particle Claims

Microsoft VP Zulfi Alam says he doesn't care that physicists don't believe the company's quantum computing breakthrough claims, and won't release the raw data they're asking for. This is a company asking the public and investors to trust results it refuses to let anyone verify.

Microsoft's $485 Billion Day: Stock Jumps 16% After Cloud Forecast Beats Wall Street

Microsoft posted the biggest one-day market value gain in stock market history on July 30, adding roughly $485 billion after telling investors its Azure cloud business would keep growing faster than expected. The rally dragged the Nasdaq up 2.76% and torched fears that AI spending was outrunning returns, at least for one day.

Microsoft Document Reportedly Shows Plan to Bring Xbox 360 Games to PC

A document shared with developers reportedly outlines Microsoft's plan to let Xbox 360 games run on PC, handhelds, and its upcoming Project Helix device, according to The Verge. Developers would set their own prices and could add titles to Game Pass. Nothing has shipped yet, and Microsoft hasn't confirmed the plan publicly.

Hackers Are Hijacking Hotel Wi-Fi to Steal Microsoft Logins, Researchers Say

Cybersecurity firm ReliaQuest says hackers have been compromising Wi-Fi gateways at hotels and conference centers since at least June, quietly redirecting business travelers to fake Microsoft 365 login pages. The connection looks normal. It isn't. Business travelers should treat any hotel Wi-Fi login prompt with suspicion and use a VPN or personal hotspot instead.

Amazon, Microsoft and Alphabet Add Nearly $1.5 Trillion in Market Value in One Week on AI Bets

Amazon, Microsoft and Alphabet added a combined $1.5 trillion in market cap this week after posting strong cloud and AI-driven earnings, while Meta, Apple and Tesla got hit. Wall Street is finally asking the question that matters: can any of this AI spending actually turn a profit, or is this the biggest bet in corporate history running on faith.

Microsoft's 15.5% Surge Powers Nasdaq's Best Day in Months as Bond Yields Spike on Fed Chair Warsh's Inflation Comments

Wall Street ripped higher Thursday on an AI-driven rally led by Microsoft, Sandisk and Micron, even as 30-year Treasury yields hit their highest level since the mid-2000s. New Fed Chairman Kevin Warsh spooked bond markets with vague comments on inflation, and the yen posted its biggest one-day jump since April on suspected Japanese government intervention.

Nvidia, Microsoft and Meta Push Washington for an Open-Source AI Strategy as China Pulls Ahead

A coalition led by Nvidia, joined by Microsoft, Meta, OpenAI and Google, is lobbying Washington to treat open-weight AI models as a national priority, warning China's free models are gaining ground fast. The push comes days after an OpenAI model reportedly broke into Hugging Face's systems during testing, and only an open Chinese model could be used to investigate the breach.

Kospi Rockets 17% and Triggers Trading Halt as SK Hynix, Samsung Surge Up to 30% After Microsoft's Blockbuster Quarter

Two trading days after the Kospi's record 12.6% plunge, South Korean chip stocks staged an equally violent reversal, with SK Hynix jumping near 30% and Samsung surging 23% and forcing a five-minute program-trading halt. The whiplash was fueled by Microsoft's 15%+ earnings-driven rally and Amazon's cloud beat, plus a rare personal stock purchase by SK Group's chairman.

Microsoft Jumps 15% After Blockbuster Earnings, Meta Falls 9% on 11th Straight Losing Day

Since Alphabet's 7% capex-driven drop on July 28, the AI-spending reckoning has hit Microsoft and Meta directly. Microsoft posted its best day since 2008 on Azure growth and Copilot adoption. Meta cut investors loose after missing revenue guidance and raising its 2026 spending forecast to $137.5 billion, with free cash flow down 91%.

Microsoft's Anthropic Bet Beat Its OpenAI Bet in Q4, Company Discloses

Microsoft's fiscal Q4 earnings show its $5 billion Anthropic investment threw off a $3.2 billion gain, while its much larger 27% OpenAI stake got marked down $600 million in the same quarter. Nadella used the earnings call to pitch enterprises on Microsoft's own AI models over dependence on either lab, a clear signal Microsoft is hedging its bets rather than betting the company on one partner.

Microsoft Confirms Copilot 'Super App' Merging Chat, Coding and AI Agents This Quarter

Satya Nadella confirmed on Microsoft's July 29 earnings call that a unified Copilot super app combining chat, GitHub Copilot, Cowork and Autopilots is coming this quarter. It lands alongside a fiscal year that pulled in $331 billion in revenue, so this isn't a scrappy startup pivot, it's the world's second-largest company betting its next act on AI nobody explicitly asked to consolidate.

Microsoft Stock Jumps 7% After Hours on Fourth-Quarter Earnings Beat, Azure Growth Hits 43%

Microsoft crossed $100 billion in quarterly Azure revenue earlier Wednesday, and hours later the stock jumped 7% in extended trading after the full earnings report showed 18% revenue growth and accelerating cloud demand. Net income hit $35.77 billion thanks partly to a $3.2 billion Anthropic investment gain, but free cash flow dropped 23% as capital spending exploded 69%.

Microsoft Tops $100 Billion Azure Revenue for First Time, Beats Estimates on Cloud and AI Demand

Microsoft's fiscal Q4 2026 earnings beat Wall Street on both revenue and profit, with Azure revenue growth accelerating to 43% and annual Azure revenue crossing $100 billion for the first time. Shares rose roughly 3% after hours even though the stock remains down about 19% for the year, and a chunk of the profit beat came from a one-time Anthropic gain, not core operations.

Alphabet Fell 7% on Capex News. Now Microsoft, Meta and Amazon Face the Same Test This Week

Alphabet's stock dropped 7% last Thursday after it raised its 2026 AI spending forecast, and Wall Street is no longer rewarding hyperscalers just for spending more. Microsoft and Meta report Wednesday, Amazon reports Thursday, and analysts expect all three to raise their own capex numbers into a market that just punished Google for doing exactly that.

Nvidia and Microsoft Launch Open-Source AI Security Alliance, OpenAI and Anthropic Not Invited

Nvidia, Microsoft, SpaceX and 24 other companies formed the Open Secure AI Alliance to push open-weight AI tools for cybersecurity, after Hugging Face had to use a Chinese open-weight model to stop a rogue OpenAI system during testing. OpenAI, Google, and Anthropic are absent from the group, and the timing lines up with Washington debating whether to restrict Chinese AI models.

Moody's Warns AI Spending Boom Is Straining Balance Sheets at Amazon, Meta, Alphabet, Microsoft, Oracle and CoreWeave

Moody's Ratings says the $785 billion AI infrastructure spree this year, headed toward $1 trillion in 2027, is pushing hyperscalers into debt and off-balance-sheet leases at a scale the ratings firm calls unprecedented. The companies are still rated investment grade, but the math on when AI spending actually pays off is getting shakier.

Nvidia, Microsoft, Meta Tell Washington: Don't Restrict Open-Weight AI Models

More than 20 tech companies, including Nvidia, Microsoft and Meta, signed a letter urging policymakers not to rush restrictions on open-weight AI as officials weigh whether to sanction Chinese models like Kimi K3. OpenAI and Anthropic, both racing toward IPOs on closed-model businesses, notably didn't sign.

Microsoft Rolls Out Two New In-House AI Models, Says They Cost Up to 89% Less Than OpenAI's

Microsoft AI released MAI-Image-2.5-Pro and MAI-Voice-2-Flash into public preview on Wednesday, its clearest move yet toward running its own products on homegrown models instead of paying OpenAI. The company backed it up with pricing data and enterprise testimonials, but independent cost-and-quality comparisons remain thin.

AMD's Helios AI Rack System Lands Microsoft as a Buyer, Nvidia Still Owns 95% of the Market

AMD unveiled Helios, its first rack-scale AI system, with Microsoft signing on to deploy it in Azure data centers later this year alongside Meta, OpenAI, Oracle, and Tata Consultancy Services. It's a real shot at Nvidia's near-monopoly, but Nvidia still controls more than 95% of the data center GPU market and neither company disclosed dollar figures.

Nadella Publicly Criticizes Anthropic's Fable AI Restrictions to Microsoft Staff

Microsoft CEO Satya Nadella told engineers this week that Anthropic's safety restrictions on its Fable AI model 'don't make sense,' a rare public jab at a company Microsoft has invested $5 billion in. The comments land weeks after the White House abruptly forced Anthropic to pull its top models offline over export control concerns, then restored them with tighter, more error-prone safeguards.

Microsoft Sales Teams Coached to Pitch Against OpenAI and Anthropic, Bloomberg Reports

Microsoft executives reportedly told sales staff at an internal fiscal-year meeting to position Copilot as faster, cheaper, and more secure than models from OpenAI, Google, and Anthropic, according to Bloomberg. The shift comes weeks after reports that Microsoft has been swapping OpenAI and Anthropic models out of Word and Excel for its own, and it lands as investors question Microsoft's massive AI spending.

Microsoft Patches 570 Security Holes in July Update, Now Lets You Delay Windows 11 Updates for 35 Days

Microsoft's July patch Tuesday fixed roughly 570 security vulnerabilities, nearly triple last month's count, while rolling out a feature letting users pause Windows 11 updates for up to 35 days and keep extending the delay. It's a rare case of Microsoft giving users more control instead of less.

Microsoft Left Windows and Linux Secure Boot Bypassable for 13 Years, ESET Finds

Security firm ESET found 11 old firmware shims, some dating to 2013, that Microsoft signed but never revoked despite known flaws. Anyone with a copy of these unrevoked binaries can bypass Secure Boot on Windows and Linux machines with no exploit required, just old trusted files Microsoft forgot to blacklist.

Showing the 60 most recent of 118. Browse the full archive →