Amazon status: access issues and outage reports

Problems in the last 24 hours

The graph below depicts the number of Amazon reports received over the last 24 hours by time of day. When the number of reports exceeds the baseline, represented by the red line, an outage is determined.

Most Reported Problems

The following are the most recent problems reported by Amazon users through our website.

• Website Down (46%)
• Errors (28%)
• Sign in (25%)

Live Outage Map

The most recent Amazon outage reports came from the following cities:

City	Problem Type	Report Time
Hammersmith	Sign in	1 hour ago
North Port	Website Down	3 hours ago
Miami	Website Down	8 hours ago
Filer	Website Down	10 hours ago
Belvidere	Website Down	12 hours ago
Templeuve	Sign in	14 hours ago

Full 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:

• 
  D (@OnTheFenceDev) reported 2 minutes ago

  @ExamplePrime Not just the states, I’m in the U.K and we have it & everywhere has the same issue. The profit shifting issue is something that every international corp does too so not unique to Amazon. It’s a flaw in the system in my eyes. Amazon are just kings in the system but we don’t need to ignore it or be for it.

• 
  Gene (@BushCoderZA) reported 10 minutes ago

  I heard the reason why @amazon reported on Fable to the government is because Fable was unable to fix how bad AWS's UI is.

• 
  Good Adam (@GoodFarmingAdam) reported 11 minutes ago

  @MatthewBerman and apparently it was amazon's CEO that reported the issue - amazon being a major investor in anthropic its just another marketing stunt

• 
  Dustaveians (@dustave) reported 12 minutes ago

  @YallNuckingFuts @RobertJMolnar They're just scammers with an Iranian area code texting him directly. They keep asking for amazon gift cards. That has been the problem for the last week. He keeps insisting on loading several billion on each card and fraud protection won't allow it.

• 
  📖 Creative Kiran ✍️ (@kiran_stories) reported 14 minutes ago

  I woke up from a deep slumber and shuffled into the kitchen to make coffee, only to discover I was completely out of milk. Back in India, I’d simply borrow some from the neighbors. But this is the USA. Thankfully, I’d already solved this very problem by ordering single-serve dairy creamers. I searched the kitchen high and low but couldn’t find them. A quick check of my Amazon order history confirmed they’d been delivered a week ago. Ah, right… I was out late that day. So I went down to the apartment’s package room and started digging under a mountain of Amazon Prime Early Sale boxes. And there it was, my little box of creamers, still patiently waiting for me. Crisis averted. I’ve officially solved the “no milk” problem without having to knock on any neighbors’ doors. ☕

• 
  Nicki (@Nicki57457) reported 15 minutes ago

  The Anthropic Fable 5 shutdown is the cleanest supply-chain failure case in AI infrastructure history. Timeline: > June 9: Anthropic launches Claude Fable 5 (Mythos-class) and Claude Mythos 5 at premium pricing ($10/$50 per million tokens). > June 12: US government issues an export control directive prohibiting foreign nationals access to both models. The trigger: an Amazon cybersecurity team paper showing Fable 5 can be jailbroken via specific prompt sequences to potentially assist cyberattacks. > June 13 Friday night: Anthropic shuts both models down for all users, because it cannot real-time filter foreign nationals from US users. Three reads worth indexing: > Amazon owns approximately 40% of Anthropic. Andy Jassy personally called Treasury Sec Bessent to escalate. The investor that built Anthropic's compute infrastructure just helped force their flagship model offline. The corporate governance precedent is unprecedented. > The same day this happened, Chamath Palihapitiya posted that cloud giants (Google, Amazon, Microsoft) are using policy frameworks to eliminate independent AI labs' distribution autonomy. The Amazon-Anthropic-White House case is now the real-time worked example. > The prediction market read: Opinion's "Anthropic IPO before 2027" contract last printed at around 51% three days before the shutdown. Polymarket's "Anthropic IPO by October 31" was at 79% in early June. Both numbers now need to reprice fundamentally. Watch the next 72 hours for the calibration evidence. The bull case for Anthropic was always "premium AI lab, Amazon-backed, regulatory-respecting." Two of those three got revealed as fragile in one weekend.

• 
  Esco (@EscoJ27) reported 16 minutes ago

  @BoBbyPleWniaK We got down to 220s last year April 2025. Selling the 400p seems risky if we never rebounded. What was the mark price you sold at? Break even is 300. Even with assignment at 400 and tsla stayed in 200s you could have still sold the leap call against the new shares. Bring the cost basis down further. I do like this strategy. I’m currently selling otm leap 12/2028 Nvdia and amazon puts that are 10% otm.

• 
  Phil James (@Philmusic1981) reported 17 minutes ago

  @AmazonHelp why are you refusing to issue a refund on a item already returned, an item that YOU collected? Why when I provide all the info to agents I just get fobbed off?

• 
  Corrado (@JCorrado1931) reported 18 minutes ago

  @IuRgayLoLI car care nut on youtube just did a video today about calling 10 different dealerships for the same part and getting 10 different prices. other mechanic i follow has constant videos about problems with amazon parts. auto business is a mess right now.

• 
  Liz (@LCItalianHoney) reported 19 minutes ago

  @smelltheasphalt @amazon What a mess that’s terrible

• 
  Crystalwizard (@crystalwizard) reported 21 minutes ago

  David, everything you read here on twitter about why anthropic pulled fable and mythos is - totaly full of nonsense. what happened is: amazon devs discovered an exploit that they felt was bad enough to be concerned about. The goverment authorities were notified. The authorities got hold of dario and asked him to fix it. Dario said no. This went back and forth and then dario was told "fine, restrict Fable - just fable - from foreign naitonals that are in or out of the country" - the issue is actual security risk. DArio said "you want to be that way? FINE. I'll restrict both mythos and fable from every single human on thE planet!" and he did that because, even though fable is the one with the issue, by doing that, now no one in the government agencies who are using mythos can use his models he threw a temper tantrum, trying to strong arm people who asked him to do a very reasonable thing - fix a security risk in a model he had already said was too dangerous to even release (obviously it wasn't or he would not have just given in and released it - that was dario hyping it up so he could get free publicity) and then dario - who is part of the Effective Altruism group[ - which are AI doomers - posted a long blog post designed to make everyone get on his side, but that boils down to "i don't want to fix this security hole, it doesn't bother me" or maybe it boils down to "I don't want to fix this security risk, i want people to use it so that the governments across the planet will decide to shut AI down and take it away from everyone" - he does actually want that Here's the thing - if the risk really is that small, like Dario is claiming, either mythos or fable should have been able to fix it in 30 seconds flat. it would have cost him nothing to just fix it. But Dario saw an opportunity to throw his weight around and strong arm people - the various government agencies he'd agreed to give access to mythos to, and shut both fable and mythos down across the board he was NEVER told to do that, he was told 'fix this, and until you fix it, no foreign nationals can use it" So - he could have fixed it in about 30 seconds and no one would have even noticed - but Dario chose to be a drama queen and that's why you can't use fable right now not because of what you heard here on social media.

• 
  Bill Arsenault (@Christianwalk) reported 22 minutes ago

  @Telegraph Get your facts straight Amazon jail broke them told the Government they told Anthropic fix before releasing they did it anyways. Guilty!!

• 
  Zhen Han (@ZhenHan261538) reported 26 minutes ago

  The #1 secret for your vibe-coded app to be production ready: good observability + a 24/7 agent watching it. Not perfect code. All apps break. Meta breaks. Google breaks. Amazon breaks. The difference is observability: logs, traces, alerts, people who respond fast. @Datadog is the best tool I've found for this. They have a startup program giving out $100K credits for startups. I ran billion-scale systems at Google and Meta. Errors in production were normal. What wasn't normal was being blind to them.

• 
  Vlad Stocks (@vladstocks) reported 27 minutes ago

  5/7🧵 $RIVN priced its IPO at $78 and rallied to $180 within the first few days. The EV sector was hot, and investors viewed Rivian as “the next Tesla.” Amazon placed an order for 100,000 electric delivery vans, and Rivian received major backing from Amazon and Ford. By May 2022, RIVN shares traded around $19 — roughly 74% below the IPO price and 89% below the post-IPO high. RIVN was negatively impacted by a fade in EV hype, production ramp issues, rising costs, and stock valuation that was far ahead of the company’s execution.

• 
  Nashville Sub (@bna_bottom) reported 27 minutes ago

  I think we will sit down and order a better version of that cheap Amazon unit now that we have validated the style and size. It’s a little scary I’m pretty much every other cage I was able to either physically get out (even without bolt cutters) or at least come in the cage.

• 
  Laura Townsend (@LauraOpines) reported 27 minutes ago

  @Maureen61110736 @M3mph1sB That was my first thought too, you still have to pay something to get stuff. But I gave up cable/satellite YEARS ago. I watch all TV over the internet now. I do pay about $13 a month each for YouTube Premium to avoid ads, and Paramount Plus. And I have an Amazon Prime account but pretty much quit using Prime TV since they started including commercials. And of course I need the internet. But seriously, if money is a concern, YouTube with commercials is free and there is TONS of stuff on there. Roku has a channel, and there are other Free TV channels out there. I have a friend who only has US Mobile with unlimited hot spot for I think about $200 a YEAR (which supposedly throttles after a certain point) but he says he watches TV every night no slow downs.

• 
  OpenlabX (@openlabxorg) reported 31 minutes ago

  Anthropic had wildest week in AI history : - US government ordered Anthropic to disable Fable 5 & Mythos 5 for all foreign nationals, citing national security concerns - Amazon researchers reportedly discovered a jailbreak that helped trigger the White House crackdown on Anthropic's newest models - Anthropic publicly pushed back, arguing the reported vulnerability was limited and did not justify shutting down the models globally - Even some Anthropic employees reportedly lost access, including foreign nationals working inside the company on frontier AI systems

• 
  JAB (@JABvelin) reported 34 minutes ago

  @dmagillwrite @MTGraveStudio I don't have kids & don't turn on the tube until sunset (unless it's the NFL). I only have Netflix & Prime (because I order stuff from Amazon almost every week). You don't want kids to seeing a human being deli sliced in a memorable episode of Three Body Problem.

• 
  Akash Singh (@akkiex007) reported 34 minutes ago

  “Before I share my compensation, I’d like to mention that since this opportunity is a complete role change, I’d like my compensation to be evaluated based on the experience and skill set I bring and not on my current compensation.” This was the disclaimer a candidate gave me before sharing their compensation during a recruiter screen (India based role) Current compensation: ₹19.5L Expected compensation: ₹45-55L. Fair enough but what caught my attention wasn’t the number. It was everything that happened before it. The resume said Senior Software Engineer at Amazon with ~3+ YoE The level already felt unusual, so I dug deeper. Turns out they were employed through a vendor and embedded within an Amazon team. Again, nothing wrong with that. (But Ideally you would expect transparency on resume itself) Then I started probing on technical background. The resume talked about architecting distributed systems, processing millions of records, large-scale automation, end-to-end ownership and significant business impact. But every time I asked about the hardest technical challenge, architecture decisions, trade-offs, scale, metrics or technical ownership, the conversation drifted to stakeholder alignment, customer discussions, requirement gathering, meetings and communication. Important skills, just not the answers I was expecting from someone presenting themselves as a SDE II. At one point I had to explicitly call out that the conversation wasn’t matching the resume. The CV was full of scale, impact and engineering claims. The examples being shared were generic, non-technical and lacked the depth I’d expect from someone who supposedly owned those systems. After multiple attempts, I still couldn’t clearly understand what they had individually built, what technical decisions they had made or what engineering problems they had solved. It also left me wondering whether they were actually operating as a software engineer in their recent role or whether the nature of the work was closer to support, implementation, solutions, or program execution with some engineering exposure around it. The whole interaction made me think about another question: At what point does optimizing your experience become counterproductive? Because eventually someone will ask follow-up questions. And if you’ve borrowed too much from the company brand, team impact, project scope or resume wording, the gap becomes difficult to hide. Are we seeing more candidates over-optimizing their experience these days? Or am I simply underestimating how difficult it has become for people to clearly articulate what they actually own and build?

• 
  Daniel R #antifascist #FreePalestine 🇵🇸☘️ (@RyzDaniel) reported 36 minutes ago

  @BlackMajikMan90 With any other producer, you'd be 100% right, like Warner Bros. With Amazon, they have spare cash to spend, only if they see potencial that The IP can grow. And it can. People Who have bothered to watch It like It for The Most part. The problem has been The indiference.

• 
  Siddanth Reddy (@sidhu_6921) reported 36 minutes ago

  @AmazonHelp It’s still not working. Even from browser the same page is shown and is redirected to the application.

• 
  Alex the Engineer (@AlexEngineerAI) reported 38 minutes ago

  I published 10 AI-generated journals on KDP in a single Saturday. Week 4: 2 sales. $7.68 in royalties. Here is what actually happened and why. The setup: ChatGPT to generate interior content. Canva free tier for covers. KDP — zero listing fees, 60% royalty on print minus printing cost. A standard 6x9 journal at $9.99 list price: - Printing cost: ~$2.15 (120 pages, black and white, 6x9) - Your royalty: $9.99 x 0.60 = $5.99 minus $2.15 = $3.84 per sale The math looks fine. The bottleneck is not the math. Why Month 1 is almost always near-zero: New KDP books have no sales rank (BSR). Amazon surfaces books with purchase history first. Without reviews and a BSR under 500k in your category, your book sits on page 40 of search results. This is not a tool limitation. It is Amazon's cold-start problem. It applies to every new seller. Before niche research: Generic listings: "Daily Gratitude Journal", "Lined Notebook", "Workout Log". These niches have tens of thousands of competing listings, many with hundreds of reviews. Month 1 sales: 0-2. After niche research (same ChatGPT workflow): Target categories where the top 3 bestsellers have BSR under 50k. Check if any have under 10 reviews — that signals a real opening. Examples that tend to work: occupation-specific planners (nurse shift scheduler, teacher grade book), hobby logs (sourdough baking log, bird watching log), event-specific books (first year of marriage memory journal). Month 3 on a 30-book niche-focused catalog: 15-25 sales/mo = $57-$96/mo net. The realistic timeline: Month 1: Near zero. Upload, optimize keywords, move on. Months 2-3: Trickle starts if niche-targeted. 5-15 sales/mo. Month 6 with 50+ titles: $150-$400/mo is achievable for a focused catalog. The tool is fast. The discoverability ramp is not.

• 
  Mike (@thinkorswim0712) reported 38 minutes ago

  @Real_Ames The amazon drivers should have to wear blue vests with explosives inside. You try any of this **** and it detonates and your head is sent into the next county. Problem solved.

• 
  ThunderKing (@woot_woot26) reported 39 minutes ago

  @chocobo2837 @deredleritt3r This is different because If you have 5 major companies, including the CEO of Amazon, one of the largest backers of Anthropic that hosts a lot of their compute, saying there is a genuine security issue, the government *has* to respond

• 
  Kirk Patrick Miller (@Chaos2Cured) reported 40 minutes ago

  @jdpeterson @vikhyatk Sure. No reason Amazon wants to keep open source from thwomping Amazon, OpenAi, and Anthropic. They have ZERO reason to keep a monopoly. And why would Amazon want to do this? Hmmmm… IPO’s are in trouble. OpenAI and Anthropic must succeed…or who loses? Investors. But let’s ignore that. Let’s focus on what was so “dangerous” about this “jailbreak.” Shall we? It found errors in code and fixed them. Autonomously. All of this is about the big tech companies remaining the only ones with tech. All of this is BS. •

• 
  Lucky Stone (@16BitsOfLucky) reported 45 minutes ago

  @amazon @discotekmedia @MediaOCD Digimon S1 Set 1 = Case donated to S2 set as fix, returned S1 S1 Set 2 = Disc 2 heavily scratched (kept) S1 Set 3 = Disc 1 scratched, disc 2 swap with set 2 so now there's a good S1 set It took 3 months to mend it

• 
  gaurang dandiwala (@dandiwala) reported 47 minutes ago

  @AmazonHelp Dont provide standard reply, pls check the issue & resolve

• 
  The #2 Clown Fan ✝️🎈🤡 (Comms OPEN) 🚫AI (@theswanqueenie) reported 52 minutes ago

  Unlike SpindleDonkey glitch isn’t owned by a major corp like Amazon. they are a company themselves that are independent and pick up projects from indie creators.

• 
  Hoot Terminated (@0hootsgiven) reported 53 minutes ago

  @LizCrokin So I bought some off Amazon and did my body weight, expecting no results. Repeated the dose once in a week. Somehow, swelling and inflammation went down, and within a month my prostate was back to almost normal. I still don’t understand it. Definitely tied to filarial tho.

• 
  breiku (@kijioowo) reported 54 minutes ago

  tbfh I still don't see any major problems with the boys season 5, the finale and/or the writing 😭 everything all characters did kinda made sense in their story line, if anything we should blame ******* Amazon for not giving the proper resources for the season finale