Amazon Outage Map

Community Discussion

Tips? Frustrations? Share them here. Useful comments include a description of the problem, city and postal code.

Beware of "support numbers" or "recovery" accounts that might be posted below. Make sure to report and downvote those comments. Avoid posting your personal information.

Amazon Issues Reports

Latest outage, problems and issue reports in social media:

• 
  BlackIntus (@Blackintus) reported 7 minutes ago

  Bill Ackman is down 10% this year. His fund trades 20% below what its stocks are worth on the open market. That means you can buy Amazon, Microsoft, Meta, and Uber today at a 20% discount through his fund. The question isn’t whether you trust Ackman. It’s whether you trust the stocks. If yes — the discount is free money.

• 
  FinancialJuice (@financialjuice) reported 10 minutes ago

  AWS: Starting June 15, network bandwidth included at no extra cost for all Amazon GameLift server instances generation 6 or later - website

• 
  Davr Yuga (@davryuga) reported 11 minutes ago

  @amazonluna Hello, good afternoon. From Spain, my experience with Amazon Luna is awful: I don't have any latency issues in the game, but I do experience FPS drops and some stuttering.

• 
  investing (@DollarCostAvg) reported 13 minutes ago

  $MU: The Textbook Definition of Margin Expansion. Micron is the best example. Micron ($MU) is what it looks like in real life. As AI demand accelerates, Micron isn’t just growing revenue—> it’s growing higher-margin revenue. HBM, high-performance DRAM, and AI infrastructure memory are becoming a larger percentage of the business, creating a powerful earnings $flywheel. Revenue Growth + Margin Expansion + Operating Leverage = Explosive Earnings Growth. Every new AI server requires dramatically more memory than traditional computing infrastructure. As demand accelerates, Micron benefits from: 🔹 Higher DRAM pricing 🔹 Massive HBM demand growth 🔹 Increasing data center mix 🔹 Stronger product pricing power 🔹 Expanding gross margins 🔹 Operating leverage at scale As revenue grows, manufacturing and operating costs are spread across a much larger revenue base. The result is that profits grow substantially faster than sales. When revenue rises and margins expand simultaneously, profits can grow dramatically faster than sales. That’s exactly what this chart is showing. By 2027, Micron is projected to generate $133 billion in profit. Let that sink in. According to these projections, Micron would generate: ✅ More profit than Meta ($103B) ✅ More profit than Amazon ($122B) ✅ Nearly the same profit as Alphabet ($135B) ✅ Within $5B of Microsoft ($138B) I can see $MU Reaching levels as high as $3000 + this yr, potentially 4,000 in 2027. They have so much cash now, they can literally Fund any new semis lines without raising capex. ✅

• 
  LiberlandPress (@PressLiberland) reported 14 minutes ago

  Artificial intelligence Why the blocking of Anthropic's AI models is so explosive+++ An instruction from the US government has ensured that the AI developer Anthropic deactivates its advanced models "Mythos 5" and "Fable 5". The decision has sparked a discussion about the access to and control of artificial intelligence. What are "Myth 5" and "Fable 5" for AI models? Both models belong to the "myth" class of the AI chatbot Claude, which is particularly powerful, called by Anthropic. In principle, "Myth 5" and "Fable 5" are the same, but they have different protection mechanisms built in. The models can, for example, take over programming tasks or analyze documents for financial controlling, for example. What are the risks? Anthropic itself had warned in the spring at the presentation of the "myth" class against exploiting the capabilities - for example, when looking for security holes in software programs. This could be abused by actors with bad intentions, for example for cyber attacks. Institutions such as the International Monetary Fund expressed concerns. The application in the economic or military field, for example, is also considered critically. Why did the US government block access? The US government had ordered Anthropic to block its latest software for foreigners. Because this is difficult to implement in the short term, Anthropic initially cut off access for all users worldwide. The US government's move was justified by security concerns, without giving further details. There is apparently a fear that the built-in protection mechanisms can be circumvented, and thus "Myth 5" and "Fable 5" could be used for illegal activities. Amazon CEO Andy Jassy has reportedly campaigned for a ban with the government with corresponding concerns. However, millions of users had already received access. It is the first intervention of this magnitude by Washington. There has been a dispute between the US government and Anthropic for a long time about the military use of AI technology. Why is the lockdown relevant for Europe? The issue of technological sovereignty is coming into focus. There is no AI provider at a comparable level in Europe. Therefore, access to the powerful models from the USA is important. Accordingly, a stronger commitment to development in Germany and Europe is also required in order to obtain even powerful AI systems with appropriate access. There have also been demands to bring companies like Anthropic, which complain about a US government intervention, to Europe.

• 
  Darin Deters (@darin_deters) reported 15 minutes ago

  Most teams shipping agentic AI have no way to enforce safety rules mid-task. Not at the start. Not at the end. Mid-task, while the agent is actually doing things. Amazon Bedrock just added an API to fix that.

• 
  Abe (@AbeyAkecheta) reported 15 minutes ago

  @AnthropicAI should blacklist organizations like @amazon. If you cause stupid problems (@ajassy) you win stupid prizes. Then they should release a new “version” of Fable and Mythos but provide NO access to the blacklist (including the US government, since they can’t play nice)

• 
  Clynton Keel (@clyntonK1) reported 16 minutes ago

  @Alexand70608635 @FennellJW Tax evasion dwarfs the entire welfare budget annually and for decades it has been ignored 'as too difficult' to fix. Amazon and other corporations evade tax. France Recouped 700 million euro from taxing such corporations. Bashing pensioners rather than the rich paying is unjust.

• 
  AFRICA IS HOME GLOBAL (@AfricaisHOME2) reported 18 minutes ago

  Nvidia launched its first investment-grade bond offering since 2021, aiming to raise 25 billion dollars across seven tranches with maturities from two to 30 years. The deal drew 85 billion dollars in investor demand, allowing Nvidia to price the longest 30 year portion at 0.65 percentage points over Treasuries after tightening from initial talk of 0.9 percentage points. Goldman Sachs, JPMorgan and Morgan Stanley are running the sale. The company said proceeds will go to general corporate purposes, including refinancing and repaying outstanding notes, and it capped the issue at 25 billion to keep spreads tight. The offering marks a shift for Nvidia, which has funded growth through cash flow and equity rather than debt during the AI boom. It last sold 5 billion dollars of bonds in June 2021. Investors see the move as a way for Nvidia to establish a liquid benchmark for its credit cost without signaling a need for capital expenditures, even as Big Tech plans to spend over 700 billion dollars on AI infrastructure this year. The bond sale follows a rally in Nvidia shares after a recent Microsoft deal and comes amid heavy AI related debt issuance from Amazon, Alphabet and Meta. Nvidia’s stock rose 3.5 percent on the news, closing at 212.45 dollars. - World Business News.

• 
  Born to gamble (@borntogambles) reported 21 minutes ago

  $10,000 in 10 days. A brand-new Shopify store. The creator says Claude AI did 90% of the work. The product nobody would guess: snake shin protectors. He started with hard criteria. Solves a problem. 4.5+ on Amazon. $35 minimum profit. Rising demand. 3 to 5 active competitors on Facebook or TikTok. Then he handed the search to Claude. He ran Claude Opus 4.8 in the desktop app's co-work mode, wired to Winning Hunter's MCP, an ad library pulling live spend data from the biggest dropshippers. He set it to act without asking. Ten minutes later Claude returned 10 products with ads dated between May 30 and June 9, each with the name, the problem, the price, the store, the competition. The reason he trusts the machine over himself: humans pick products with their feelings. The creator says beginners drown in options and freeze. Claude doesn't freeze. It chose the snake gaiters off the data and named the buyers: hikers, dog walkers, people scared of snakes. A small niche older buyers shop, where AI video ads read as real because the audience doesn't know what AI looks like yet. He cloned a winning Facebook ad inside Higgsfield's marketing studio, generated an avatar from a prompt (a gray-haired man in his 50s who looks like he works outdoors), pulled a voice off ElevenLabs, and assembled the spot in CapCut. The original creative cost a studio. His cost a prompt. Facebook strategy on a Miro board: $50/day, 3 broad ad sets, 3 ads each, 9 total. His dashboard from April 26 to May 11: gross sales over $11,000. Average order value $47. Conversion rate 3.6%, above industry standard. By June 8 the store was clearing roughly $1,000 a day. A product 99% of beginners would never touch made $11,000 because a machine had no taste to get in the way.

• 
  Karp (@Karpeltunnel) reported 24 minutes ago

  @TheCryptoDaddi The reason I went with Amazon is they accept insurance so I’m paying $8/mo. I just lied and said that I had used daily tadalafil before for ED and it worked without issue.

• 
  Duane - 🧙‍♂️🖖 - keybase.io/dfk (@honestduane) reported 26 minutes ago

  @QuinnyPig @ajassy Even a broken clock is right twice a day. Amazon and anthropic compete with each other. It makes sense that Amazon would want to sabotage a competitor that’s doing better than them, has a better reputation than them, etc. Amazon‘s problem is that it’s not customer focused enough.

• 
  Marcus (@Marcus70672192) reported 27 minutes ago

  I noticed yesterday that all the Indian takeaways next to me- of which there many- have bought a load of fake reviews to give themselves a 5-star rating on Just Eat. Apparently there is an enquiry into this taking place as a whole, which will be settled in September. But that's terribly unethical I think. It's also a big problem on Amazon. So it's getting much harder to determine product quality across the board. And impossible, pretty much, on just eat.

• 
  H Hart (@thecai_bleu) reported 29 minutes ago

  @1LovedPuppers FOR THE LOVE OF HUMANITY RESEARCH THE LADY WHO CREATED THE MASKS , JONNA MENDEZ. I gave you all that information yesterday. PURCHASE HER BOOK ON AMAZON. Type in mask on X. The military sees all my posts. I could get in serious trouble typing lies. RESEARCH HER. Masks are being used on hundreds of people helping the military take down the Illuminati Deep State Freemasons Cabal. Watch Boston Blue Episode 19 on CBS. They did a mask thing. I do not know how to help you anymore than that.

• 
  Penny Portrait (@PennyPortrait) reported 30 minutes ago

  @amznsellerhelp The stupid Amazon AI system has delisted my product 50+ days because the description mentioned glue. It decided to flag my poster as a hazardous material. Every time I reach out I get a new person from the Philippines who "escalates" my issue. Amazon doesn't care.