Paul retweeted
Wow.
התנהלות ראש הממשלה היוצא נתניהו וסביבתו היא איום מוחשי, ישיר ומידי על הדמוקרטיה הישראלית.
לא נותרה בישראל חלקה שלא הושחתה, עוותה והוכפפה למען הישרדותו הפוליטית האישית של נתניהו. אין רסן שלא איבד, ואין גבול שלא נחצה.
אסור בשום פנים ואופן שנתניהו ימשיך לאחוז בהגה השלטון ולו דקה אחת נוספת. אסור בשום פנים ואופן שישראל תיגרר לסבב בחירות נוסף כשהוא ראש ממשלה. מובן שכל עוד הוא בתפקיד - לא יתקיימו יותר בחירות הוגנות וכשרות בישראל. זה השיקול המרכזי שמנחה אותי בכל משא ומתן קואליציוני שאני מנהל בימים אלו.
לאחר שהכניס ללשכת ראש הממשלה סוכנים קטארים בפרשת "קטארגייט", עכשיו נחשפת פרשיה חדשה וחמורה עוד יותר - ניסיון שיבוש בחירות לכאורה תוך השתלת ראיות, הפעלת המשטרה נגד אזרחי ישראל הערבים בלי צו חיפוש או צו מעצר בתוקף.
אני קורא לפתיחת חקירה מידית של הפרשה ולהבהרה מידית של המשטרה והשב״כ בדבר הפרטים שפורסמו.
אפעל להקים ועדת חקירה לחקר מעורבות והתערבות לשכת ראש הממשלה ביום הבחירות והתנהלות גופי הביטחון לקראתו ובמהלכו.
Paul retweeted
🦔 Nvidia’s days as the AI industry’s open checkbook look numbered, according to The Wall Street Journal. Interim CEO Colette Kress is expected to rein in the company’s financing deals, and UBS estimates a pullback could take $50 billion or more out of the AI market.
My Take
For the past year Nvidia acted like the AI boom’s lender of last resort. It took part in 129 venture deals this year, worth about 27% of all the venture money raised worldwide by PitchBook’s count, which is wild for a company that makes chips. Whenever a cloud startup or a lab ran short, Nvidia’s cash or its guarantee showed up. That’s ending, and the money has to come from somewhere else.
Private credit already charges outfits like CoreWeave junk-bond rates, so that door keeps getting more expensive. That leaves the stock market. Anthropic goes public this week with about a fifth of its shares set aside for regular investors, and OpenAI wants to follow early next year. The risk Nvidia carried for the last two years is about to shift into IPOs and 401(k)s, and I advise anyone to think hard before buying any of these on day one.
Hedgie🤗
Paul retweeted
'You do it overnight when there’s no traffic'
1) work in a global org and find out what overnight means...
2) airplanes don't have time off
3) sometimes you should 100% patch during the day
it's way more complicated than this person is making out.
This is a lie. You do not, I repeat, do not EVER do a system update/upgrade at 3pm on a busy day. You do it overnight when there’s no traffic. That goes for any type of system update. This is complete bullshit. Here’s what probably actually happened: this new AI system went live on November 10th, overwhelmed the existing infrastructure and crashed spectacularly on November 13th.
What I find incredible, and am thankful for, is that processes and logs exist to determine root cause; and to be open with us about it.
Kudos to Microsoft and all members of the team that have had a sleepless few days,
More details on the root cause can be found in the Microsoft 365 Admin center, and an abridged version can be found below.
We are deeply sorry for the impact to all our customers and recognize the impact that this has on their normal business & operations.
The incident on Thursday, November 12th, 2026, was caused by three main factors that together caused a global outage for the networking gateway that fronts many Microsoft 365 services.
First, a latent defect caused servers to become overwhelmed and unresponsive under certain load conditions. This was due to lock contention inside an internal process which caused the machine to become unavailable. This defect was the result of a fix that was prepared specifically for Microsoft 365 networking gateway servers; neither the underlying defect nor the fix applies to broader Windows deployments.
The initial change in behavior first started deployment in July of this year. The update went through safe deployment and saturated globally over several weeks, during which time no issues were observed. For more than three months, the change was deployed worldwide without impact.
Our normal load testing did not detect the latent issue because the different traffic and SKU mix used in our test labs. The contention issue is worsened on machines with more cores and newer hardware; whereas our load testing tended to focus on our weaker hardware SKUs with fewer cores.
Second, a surge in load increased the likelihood for the networking gateways to encounter the contention. Within Microsoft 365, we maintain two independent and heterogenous network gateway systems designed for isolation and high availability. This allows many Microsoft 365 services to fail to/from them during outages or disruptions. Approximately fifteen hours before the outage, Teams failed over from the alternative networking gateway to the ‘primary’ gateway that also supports Exchange Online. This increased the load on the networking gateway and changed its traffic patterns. This type of failover is part of normal operations and is integral to standard Microsoft resiliency efforts.
Third, at the same time, nearly 14% of server capacity for the networking gateway was out of service due to an unrelated issue from the previous day. This server capacity was on track to be returned to active service through routine maintenance and health processes. The service had sufficient capacity headroom, so the reduction in server capacity had no impact on its own.
Taken together, the increased load and the reduced server capacity, caused the network gateway to encounter the latent bug from July. This caused requests to build up until servers ultimately stopped responding. This occurred during the peak window of traffic in the United States.
When servers lose connectivity, our traffic management systems automatically route users to other healthy endpoints. This happens often and is critical to avoiding any interruption due to hardware, software, or network defects and is normally what makes the service resilient. However, the redirected traffic increased the likelihood for the same problem to occur on the remaining endpoints. This caused a cascading failure from server to server as the same amount of load was spread across fewer servers. In less than two hours, the previously unknown defect was encountered across our fleet and rendered them unresponsive. At its lowest, more than 80% of servers were unresponsive.
Normally, the network gateway is composed of thousands of ‘capacity units’ which represent shards of capacity, hardware, and networking infrastructure. These ‘capacity units’ are then organized into regions. In this case, due to a separate issue, the normal bulkheads between these different partitions did not stop the cascading failure. This is what allowed the impact to spread across our network gateway infrastructure globally.
Several services like Teams, OneDrive, SharePoint, and Office apps failed out of the faulty network gateway infrastructure. That is why those services were restored earlier, with Teams chats and meetings functioning without issue for most of the outage. But not all services, like Exchange Online, can fail out of this network gateway.
Recovery took longer than usual because as servers returned to service, they were immediately overwhelmed by waiting traffic and became unresponsive again. Our normally automated recovery processes were not designed to restore service from this broad unresponsive state. Additionally, since the main code defect was a latent bug from several months previous, our normal rollback, flighting, and feature flags were unable to quickly revert the change.
After engineers confirmed the defect as the cause, we developed an update that returned the code path to the previously known good state. To effectively apply this fix, and to slowly bring the service online again, we took groups of servers out of service, applied the change, and restarted them before returning them to service. This process was carefully applied to large quantities of infrastructure at a time (e.g. 50%+ of a region).
The final Post-Incident Review (PIR) will provide more details later this week.
49
Paul retweeted
More details on the root cause can be found in the Microsoft 365 Admin center, and an abridged version can be found below.
We are deeply sorry for the impact to all our customers and recognize the impact that this has on their normal business & operations.
The incident on Thursday, November 12th, 2026, was caused by three main factors that together caused a global outage for the networking gateway that fronts many Microsoft 365 services.
First, a latent defect caused servers to become overwhelmed and unresponsive under certain load conditions. This was due to lock contention inside an internal process which caused the machine to become unavailable. This defect was the result of a fix that was prepared specifically for Microsoft 365 networking gateway servers; neither the underlying defect nor the fix applies to broader Windows deployments.
The initial change in behavior first started deployment in July of this year. The update went through safe deployment and saturated globally over several weeks, during which time no issues were observed. For more than three months, the change was deployed worldwide without impact.
Our normal load testing did not detect the latent issue because the different traffic and SKU mix used in our test labs. The contention issue is worsened on machines with more cores and newer hardware; whereas our load testing tended to focus on our weaker hardware SKUs with fewer cores.
Second, a surge in load increased the likelihood for the networking gateways to encounter the contention. Within Microsoft 365, we maintain two independent and heterogenous network gateway systems designed for isolation and high availability. This allows many Microsoft 365 services to fail to/from them during outages or disruptions. Approximately fifteen hours before the outage, Teams failed over from the alternative networking gateway to the ‘primary’ gateway that also supports Exchange Online. This increased the load on the networking gateway and changed its traffic patterns. This type of failover is part of normal operations and is integral to standard Microsoft resiliency efforts.
Third, at the same time, nearly 14% of server capacity for the networking gateway was out of service due to an unrelated issue from the previous day. This server capacity was on track to be returned to active service through routine maintenance and health processes. The service had sufficient capacity headroom, so the reduction in server capacity had no impact on its own.
Taken together, the increased load and the reduced server capacity, caused the network gateway to encounter the latent bug from July. This caused requests to build up until servers ultimately stopped responding. This occurred during the peak window of traffic in the United States.
When servers lose connectivity, our traffic management systems automatically route users to other healthy endpoints. This happens often and is critical to avoiding any interruption due to hardware, software, or network defects and is normally what makes the service resilient. However, the redirected traffic increased the likelihood for the same problem to occur on the remaining endpoints. This caused a cascading failure from server to server as the same amount of load was spread across fewer servers. In less than two hours, the previously unknown defect was encountered across our fleet and rendered them unresponsive. At its lowest, more than 80% of servers were unresponsive.
Normally, the network gateway is composed of thousands of ‘capacity units’ which represent shards of capacity, hardware, and networking infrastructure. These ‘capacity units’ are then organized into regions. In this case, due to a separate issue, the normal bulkheads between these different partitions did not stop the cascading failure. This is what allowed the impact to spread across our network gateway infrastructure globally.
Several services like Teams, OneDrive, SharePoint, and Office apps failed out of the faulty network gateway infrastructure. That is why those services were restored earlier, with Teams chats and meetings functioning without issue for most of the outage. But not all services, like Exchange Online, can fail out of this network gateway.
Recovery took longer than usual because as servers returned to service, they were immediately overwhelmed by waiting traffic and became unresponsive again. Our normally automated recovery processes were not designed to restore service from this broad unresponsive state. Additionally, since the main code defect was a latent bug from several months previous, our normal rollback, flighting, and feature flags were unable to quickly revert the change.
After engineers confirmed the defect as the cause, we developed an update that returned the code path to the previously known good state. To effectively apply this fix, and to slowly bring the service online again, we took groups of servers out of service, applied the change, and restarted them before returning them to service. This process was carefully applied to large quantities of infrastructure at a time (e.g. 50%+ of a region).
The final Post-Incident Review (PIR) will provide more details later this week.
Replying to @MSFT365Status
We have released the Preliminary Post-Incident Report of last week’s outage, which is accessible via the Admin center under MO1550821 and EX1550543.
Paul retweeted
FFMPEG had one of the most valid crash outs in infosec history IMO. There's a huge elephant in the room when it comes to the fact that cybersecurity is extremely unbalanced. You have like 1,000 people pointing at problems for every 1 person fixing them.
Vulnerability research is extremely important, but it's become somewhat of a marketing stunt. First with fuzzers, and now with LLMs. It's a lot of "look what my new toy can do" rather than earnest attempts at actually fixing security problem.
What Google did to open source developers was especially egregious. They're a $3.6 trillion dollar company, who spent billions upon billions of dollars building a bot that points at problems really really fast, then they just dumped them all on a bunch of unpaid open source project maintainers to fix.
Then because of Google's 90 day disclosure policy, the maintainers have 90 days to fix the issues. After 90 days, Google goes public with the issues, making their project look bad if the vulnerability is yet to be fixed, and also potentially harming users if it gets exploited.
The 90 day policy was intended to give developers a reasonable amount of time to fix issues, while also applying some pressure with the threat of disclosure. Otherwise, you do get a bunch of goober ass vendors that will just ignore all security issues reported to them, then act completely surprised when their software gets hacked.
But when you're a small team, non-profit, or a solo developer, and you have a trillion dollar company's AI just blasting you with infinite problems to fix (each coming with a 90 countdown & a threat), that becomes extremely untenable. They're essentially inadvertently blackmailing developers into becoming unpaid interns.
Then on top of that, Google profits unimaginable amounts of money from the use of open source code. They use FFMPEG's code extensively across several highly-profitable products, meanwhile FFMPEG's developers work for free. There's no reason Google couldn't have donated money, engineers, AI tokens, or something towards fixing the problems. They can more than afford it.
I don't think Google has any malicious intent at all, it's just that the whole incentive structure behind vulnerability research is completely whack. All the glory goes to people who find issues, not those who fix them, so there's not a whole lot of interest in helping with fixes. Some vulnerability researchers will submit fixes with their reports, but most don't.
Honestly, in 2026 cybersecurity basically just feels like the equivalent of 50,000 people following the housecleaner around yelling "you missed a spot". Abysmally little effort has gone into anything outside of pointing out vulnerabilities.
Paul retweeted
A lot of people say: “But Nvidia’s stock fell because there was an extramarital affair. In fact, there was more than one, and the main one was with a Chinese spy.”
Come on… the fact that she only now appears as a Chinese spy just makes me believe there is government involvement in this.
They attended countless events together and sometimes even held hands. Only now, after the offensive against Jensen began, does the accusation that she is a spy surface?
Then these same people say: “But extramarital affairs with subordinates inside companies end in dismissal.”
They only end that way when powerful people want it. Normally these cases are suppressed well before they gain any real traction.
In the case of a company like Nvidia, neither the company itself nor the government would have any interest in letting this scandal explode. Everything is normally done to protect the corporation, especially one of Nvidia’s size and strategic importance.
Unless an operation was already in place to expose Jensen’s personal life and directly link it to a national security risk: a multi-front, well-orchestrated attack.
Once the national security angle is introduced, the matter triggers an investigation and severely weakens the CEO’s position to remain in the role.
Paul retweeted
Change management …. Easy to say you do it, much harder to do it well!
Planning, by and large, helps to mitigate not prevent.
I’m very curious about _how_ incidents like the Microsoft outage and FAA ground stop were allowed to happen in the first place.
In other words, what were the decisions made prior to those outages.
Not because there is necessarily causality there (stuff happens) but by analyzing our past decisions and the eventual outcomes I believe it helps us plan and prepare better for future situations.
Paul retweeted
It's time to panic about AI.
For years, I listened to the reasonable people. I nodded along when critics called AI “a con,” and when they waved ChatGPT away as “spicy autocomplete.” I laughed in 2024, when the internet was mocking ChatGPT because it couldn’t count the r’s in “strawberry.” When I met young researchers who had devoted their careers to the danger of rogue AI, I thought: what a waste of talent.
But at some point, I started to realize something was off. The machine kept learning things it was never supposed to learn.
OpenAI recently published more than 700 papers with the solutions to hundreds of open problems in mathematics. Hugo Duminil-Copin, winner of the Fields Medal (the highest prize in mathematics), wrote: “The shock is immense. I am paralyzed.”
None of this should have happened. So I went back and looked at what else the skeptics had predicted, and it dawned on me: they had been wrong about... pretty much everything. Next, I looked at the predictions of the crazy people, and realized they’ve been eerily prescient.
Why then, you may wonder, does the world still feel pretty normal? Because new technology usually spreads more slowly than its inventors expect (it took factories decades to reorganize around electricity). But the scary scenario is not about what happens in the wider economy; it’s about what happens inside a handful of companies.
I wrote down what changed my mind, and what we can do about it:
rutgerbregman.substack.com/p…
Paul retweeted
IMF가 AI 판에 얽힌 순환출자(Circular Financing) 거래 규모가 무려 1.4조 달러(한화 약 1,900조 원)에 달한다고 공식 경고장 날림 ㅎㄷㄷ 🤯
한 줄로 요약하면 빅테크랑 칩 메이커들이 자기네 물건 사줄 고객사한테 투자하고 빚보증 서준 다음, 그 돈이 돌고 돌아 다시 빅테크 매출로 찍히는 거대한 돌려막기 구조임 ㅋㅋ
지금 시장에 공개된 AI 순환 거래 17개 싹 다 정리함. 숫자 규모가 진짜 정신 나갔음.
1. 엔비디아 $NVDA ↔ 오픈AI
엔비디아가 오픈AI에 300억 달러 투자하고, 오하이오 데이터센터 리스도 최대 1,050억 달러 보증 서줌. 오픈AI는 그 데이터센터를 엔비디아 칩으로 꽉꽉 채움.
2. 오라클 $ORCL ↔ 오픈AI
오픈AI가 오라클이랑 3,000억 달러짜리 클라우드 계약 맺음. 오라클은 이거 지어주려고 영끌해서 빚내는 중임.
3. 마이크로소프트 $MSFT ↔ 오픈AI
마소가 오픈AI 지분 약 27% 들고 있고, 오픈AI는 마소 애저(Azure) 클라우드에 2,500억 달러 쓰기로 약정함.
4. 소프트뱅크 ↔ 오픈AI
소프트뱅크가 오픈AI에 646억 달러 때려 박았고, 오픈AI는 소프트뱅크 자회사 SB에너지 데이터센터 빌려 씀.
5. 아마존 $AMZN ↔ 오픈AI
아마존이 오픈AI에 최대 500억 달러 투자 중이고, 오픈AI는 아마존 클라우드 계약을 1,000억 달러 추가 확대함.
6. AMD $AMD ↔ 오픈AI
AMD가 오픈AI한테 자사 주식 최대 1.6억 주를 주당 0.01달러(사실상 공짜)에 살 수 있는 워런트 줌. 대가로 오픈AI는 AMD 칩 사줌.
7. 구글 $GOOGL ↔ 앤스로픽
구글이 앤스로픽에 최대 400억 달러 투자하고, 앤스로픽은 구글 클라우드에 최소 1,210억 달러 쓰기로 약정함.
8. 아마존 ↔ 앤스로픽
아마존이 앤스로픽에 최대 330억 달러 투자하고, 앤스로픽은 AWS에 최소 1,100억 달러 약정함.
9. 브로드컴 $AVGO ↔ 앤스로픽
은행들이 특수목적법인(SPC)에 630억 달러 대출해줘서 브로드컴 칩 사다가 앤스로픽에 리스해줌. 브로드컴은 그 칩 중고 잔존가치까지 보증 서줌.
10. 마소 + 엔비디아 ↔ 앤스로픽
둘이 합쳐 최대 150억 달러 투자하고, 앤스로픽은 마소 클라우드에 최소 449억 달러 약정함.
11. 스페이스X $SPCX ↔ 앤스로픽
스페이스X가 400억 달러 대출 당겨서 엔비디아 칩 사고, 그 컴퓨팅 파워를 앤스로픽에 팜. 참고로 엔비디아는 스페이스X 주주임.
12. 엔비디아 ↔ 코어위브 $CRWV
엔비디아가 코어위브 지분 약 11% 들고 있고, 안 팔리고 남는 용량 63억 달러어치 사주기로 보장함. 코어위브는 당연히 엔비디아 칩 삼.
13. 엔비디아 ↔ 네오클라우드
엔비디아가 중소형 AI 클라우드 업체들한테서 360억 달러어치 용량 사주기로 약속함. 얘네는 그 확정 매출 담보로 대출받아서 엔비디아 칩 삼.
14. 마소 ↔ 엔스케일, 네비우스 $NBIS, 아이렌 $IREN
마소가 얘네랑 최대 709억 달러 계약 맺어주고, 얘네는 그 돈 받아서 GPU 사들임.
15. 구글 ↔ 전직 비트코인 채굴업체들
구글이 테라울프 $WULF, 사이퍼마이닝 $CIFR 데이터센터 리스 수십억 달러 보증 서주고 대가로 얘네 지분 챙김.
16. 메타 $META ↔ 블루아울 $OWL
합작법인으로 273억 달러짜리 데이터센터 짓고, 메타가 280억 달러 규모 보증으로 뒷배 서줌.
17. 앤스로픽 IPO
칩·클라우드 공급사인 엔비디아, 구글, AMD, 마이크론 $MU, 브로드컴, 마소가 앤스로픽 IPO 지분 260억 달러어치 사줌. 정작 앤스로픽은 이 회사들한테 앞으로 내야 할 돈만 수천억 달러임 😂
이게 왜 중요하냐면, 돈이 계속 돌고 빚이 굴러가는 동안에는 빅테크부터 반도체, 데이터센터까지 전부 분기마다 어닝 서프라이즈 찍으면서 매출이 미친 듯이 떡상하는 것처럼 보임.
근데 핵심은 이거임. 이 1.4조 달러짜리 파티는 실제 최종 소비자들이 AI 서비스에 돈을 펑펑 써줘야 유지되는 구조임. 만약 최종단 수익화가 기대보다 삐끗하는 순간, 밑단에서 빚내서 케펙스(CAPEX) 돌리던 AI 모델사·네오클라우드부터 터지고, 지분 투자에 리스 보증까지 얽힌 빅테크들까지 연쇄적으로 뚝배기 깨질 수 있다는 게 IMF의 팩트 폭격임.
결론: 돈 돌 때는 세상에서 제일 화려한 영구기관인데, 펌프 멈추는 순간 역대급 도미노로 변할 수 있으니 매출 숫자 뒤에 숨은 보증이랑 부채 규모는 꼭 체크해야 함.
※ 특정 종목 매수·매도 추천이나 투자 자문 아님. 판단과 책임은 본인에게 (Not Financial Advice)
Microsoft is spending $175 billion on AI while offering buyouts to 60,000 employees.
17 days later, a routine network update took down Outlook, Teams, and Copilot across 74 countries.
430 million users were locked out for 18 hours while European hospitals went back to paper fax machines.
This is the story of incident EX1550543 and how Wall Street's favorite AI accounting trick broke the internet's backbone.
October 26, 2026
Microsoft Chief People Officer Amy Coleman sends an internal memo across the US, UK, Canada, and India.
The company is rolling out a Voluntary Exit Offer to 60,000 employees, roughly 27% of its global workforce.
The buyouts reach all the way up to Level 70:
- veteran systems engineers
- senior directors and general managers
- institutional memory built over 20 years
Why would a profitable tech giant push 60,000 people toward the door by December 31?
Because of how Wall Street grades AI spending:
- $175B in AI server racks is CapEx, depreciated slowly over years
- 60,000 human salaries are OpEx, deducted 100% from quarterly margins right now
Cut the people, buy the GPUs, and your operating margin looks like a masterpiece.
November 12, 2026
17 days after the buyout memo, at 12:30 PM EST, a network config change hits Exchange Online.
A large batch of core servers suddenly fails Microsoft's automated health checks.
The automated system does what it was programmed to do and pulls every failing server offline at once.
Then the real nightmare starts.
Live traffic from 430 million users dumps onto the few remaining servers in seconds.
The sudden flood crushes the login and routing layer, sending 140,000 outage reports straight up on Downdetector.
Outlook goes down. Teams goes down. SharePoint goes down.
Even Copilot, the flagship AI assistant meant to replace human workflows, goes dark because it runs on the exact same legacy plumbing.
Every time engineers try to restart a dropped server, the wall of waiting traffic immediately freezes it again.
18 hours later
No AI agent could break the loop.
Human engineers had to manually halt deployments, override the health checks, and bleed traffic back in slow stages while extending email timeouts to 48 hours.
To this day, Microsoft incident codes EX1550543 and MO1550821 stand as a live stress test of replacing institutional memory with automation.
You can spend $175 billion on AI clusters and cut 60,000 roles to make a quarterly earnings slide glow for Wall Street.
In enterprise infrastructure, senior engineers aren't bloat on a spreadsheet, they are the only circuit breaker standing between a bad config file and a global blackout.
Paul retweeted
🦔In just over a day, Microsoft’s email went dark for 13 hours and the FAA grounded every departing flight in the country until it lifted the order tonight. Officials say neither was a hack, and both appear to trace back to routine updates that went wrong.
My Take
Hackers spend years trying to shut down something as big as U.S. air travel, and a botched update did it by accident. Flights are moving again, and I’d treat today as a free warning about where the weak spots are.
We spent the last decade squeezing the slack out of everything. Companies cut backup systems and experienced staff to save money, and now AI writes a growing share of the code (Microsoft’s CEO has said up to 30% in some projects). It all looks great in an earnings report until the day it breaks, and then the bill falls on travelers and workers instead of the companies.
Banks got stress tests after 2008 because one failure could drag everyone down with it. So much of daily life now runs through a handful of tech and telecom companies that they deserve the same kind of test, and big customers and shareholders can start asking for proof these systems survive a bad day before a bigger outage forces the question.
Hedgie🤗
Paul retweeted
My Manager Got a Bonus for Cutting Mine. Funny how that works, isn't it? 😂
🤖 Made with AI
Paul retweeted
We are deeply sorry for the disruption this incident has caused our customers and partners. We know the trust they place in our services and take that responsibility extremely seriously. We are seeing broad recovery and stability across our services.
This was a technical issue not related to cyber or AI / agents. The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change. We have addressed the underlying issue and continue to restore service to remaining customers and regions.
Given the scale of demand, we restored services gradually, validating stability at each stage. While this took longer, it helped ensure a stable restoration.
We will provide a full post-incident report once our review is complete.
It was internal DNS wasn't it.
It's always bloody DNS 🤷♂️😂
We are deeply sorry for the disruption this incident has caused our customers and partners. We know the trust they place in our services and take that responsibility extremely seriously. We are seeing broad recovery and stability across our services.
This was a technical issue not related to cyber or AI / agents. The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change. We have addressed the underlying issue and continue to restore service to remaining customers and regions.
Given the scale of demand, we restored services gradually, validating stability at each stage. While this took longer, it helped ensure a stable restoration.
We will provide a full post-incident report once our review is complete.
Paul retweeted
Thanks to everyone who worked to resolve this! It doesn’t excuse the incident but the efforts are appreciated! ❤️🖖
Incidents come in all shapes and sizes, some are dumb, some are incredibly complex and rare….
But in the end the people that work to fix them… well they are appreciated!
We are deeply sorry for the disruption this incident has caused our customers and partners. We know the trust they place in our services and take that responsibility extremely seriously. We are seeing broad recovery and stability across our services.
This was a technical issue not related to cyber or AI / agents. The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change. We have addressed the underlying issue and continue to restore service to remaining customers and regions.
Given the scale of demand, we restored services gradually, validating stability at each stage. While this took longer, it helped ensure a stable restoration.
We will provide a full post-incident report once our review is complete.




























