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 (47%)
• Errors (28%)
• Sign in (24%)

Live Outage Map

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

City	Problem Type	Report Time
Petaluma	Sign in	4 hours ago
Hartford	Errors	4 hours ago
Newnan	Errors	20 hours ago
Ashburn	Sign in	1 day ago
North Las Vegas	Sign in	1 day ago
Saint-André-de-Corcy	Sign in	2 days 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:

• 
  Ankit Sharma (@Ayush28065779) reported 7 minutes ago

  @AmazonHelp If you want, I can delete the tweet containing the tracking ID. But how can I be sure my issue will actually be addressed after I delete it? I need assurance that my case is being worked on and my delivery will be expedited. I have an important interview on Tuesday,

• 
  Satguy 141 (@satguy01) reported 8 minutes ago

  Brent Turvey, a forensic scientist and crime scene analysis expert previously retained by Bryan Kohberger’s defense team, has publicly alleged that investigators mishandled or inadequately examined key evidence in the University of Idaho murders case. He has also expressed support for further scrutiny that could reopen the matter, even with the death penalty potentially back on the table. Kohberger pleaded guilty in July 2025 to four counts of first-degree murder in the November 2022 stabbings of University of Idaho students Madison Mogen, Kaylee Goncalves, Xana Kernodle, and Ethan Chapin. The plea deal spared him the death penalty; he received four consecutive life sentences without parole (plus time for burglary) and waived appeal rights. Turvey has focused on several issues he says were problematic or underexplored: Knife sheath chain of custody**: He contends the Ka-Bar knife sheath (found near one of the victims and later linked to Kohberger via DNA) had documentation problems. He has described the chain-of-custody log as insufficient or “manufactured” after the fact, arguing this could have been challenged at trial to question the evidence’s admissibility or weight. Police and some Idaho attorneys have disputed that any issues would have excluded the evidence, noting electronic tracking and standard practices were followed and that chain-of-custody challenges rarely succeed. Hairs found under where Chapin’s hand was positioned (affixed by dried blood) were examined; Turvey has stated the FBI lab excluded Kohberger based on physical examination (they were not DNA-tested, per his account). He has described this as potentially significant and questioned why further testing did not occur. Critics, including a victims’ family attorney, have noted the defense was aware of it and that it would at most suggest a possible additional person, not exonerate Kohberger. Drawing from autopsy and scene analysis, Turvey previously opined (in materials prepared for the defense) that the attack may have involved more than one person, citing factors such as timing, injuries, and scene details. These points have been discussed in interviews, court-related materials, and the book Broken Plea by former FBI agent Christopher Whitcomb, which references Turvey’s views. Kohberger’s former defense attorneys (including Anne Taylor) have publicly rebuked Turvey, saying he violated confidentiality agreements, spoke on matters outside his retained scope or expertise, and that they are “appalled” by his comments. Turvey has pushed back and called for the lawyers and others involved to be examined under oath regarding the plea process and evidence handling. Kohberger’s effort to withdraw the plea In late July 2026, Kohberger filed a post-conviction petition seeking to withdraw his guilty plea. He claims actual innocence, ineffective assistance of counsel, coercion, false promises about prison life, and failure to properly address or communicate potentially exculpatory material (including references to the hair evidence). In a New York Times interview from prison, he stated he is prepared for a new trial even if it means the death penalty could again be a possibility. Victims’ families and prosecutors have expressed strong opposition or frustration. Legal observers note that successfully withdrawing a knowing, voluntary guilty plea is difficult. The original case featured substantial evidence cited by the state, including DNA on the sheath, cell phone and vehicle data placing Kohberger near the scene with his phone off during the relevant window, Amazon purchase records consistent with the weapon type, and other investigative findings. Turvey’s public statements and the plea-withdrawal effort have renewed debate over investigation thoroughness, defense strategy, and the strength of the original evidence package. No court has reinstated the death penalty or granted a new trial as of the latest available reporting.

• 
  Jon Cohen (@TheJonCohen) reported 9 minutes ago

  @BestBuySupport @BestBuy lets have this conversation publicly. why did you run me around your entire store, customer service, geek squad, each person with a different answer, asked for manager- was told he was on break, then told i should try online help. You know who solved my problem, Amazon

• 
  Vinita Sharma (@Vinita_Sharma30) reported 11 minutes ago

  @AmazonHelp The link was not working.

• 
  Siddharth Raja (@rajkotraja99) reported 14 minutes ago

  @AmazonHelp Not working

• 
  Rahul Yadav (@r_yadav_26) reported 15 minutes ago

  @AmazonHelp @amazonIN Firstly, they charged me rs 300 visiting in 4 months for atleast 3 times. I still have a problem. And they will charge me again. What's the point of warranty secured by amazon. Being a loyal customer for nearly 18 years in uk and india. I am waisting time and mone

• 
  Luke Pierce (@lukepierceops) reported 21 minutes ago

  Amazon is putting $1 billion into hiring thousands of engineers to go sit inside other companies and build AI systems for them. Friday the stock had its best day since 2012. Up 15.3%, AWS growing 37%. And this is the clearest signal yet that implementation is the business. Anthropic launched Ode two weeks ago. OpenAI stood up the Deployment Co. before that. Palantir has been running the model for over a decade. Amazon is the biggest cloud provider on the planet and the first hyperscaler to formalize it, and they're doing it while they've cut 30,000 corporate roles since 2022. We've been running this exact model at Boom since before it had a name, 87 clients in. And this is the opportunity right now. The work is walking into a company, mapping how it actually operates, consolidating everything into one system, and layering AI on top of processes that hold. Priced on outcomes instead of hours. Weeks instead of quarters. That's the whole business. There's a lot of noise around this right now. Agencies selling agents to companies whose underlying processes are broken. You can drop an agent into chaos and all you get is faster chaos. The adoption rates there are less than 50%. The other thing that people are missing here is the partner motion. AWS is credentialing outside firms to deliver this work under their own engineering bar. Invite-only right now while they validate it, opening up after. A company with thousands of engineers and a billion dollars in budget is out looking for help because the demand is bigger than what they can staff. The accounts they're taking themselves look like what you'd expect. The NFL, Cox Automotive, The Allen Institute, etc. Meanwhile there are hundreds of thousands of companies doing $2M to $50M a year running on 20 disconnected tools with full access to the best AI ever built and no idea how to apply it. Nobody at AWS is calling them and nobody at Ode is either. Amazon is showing you what every company is going to be buying over the next several years. Go be the one who builds it for them.

• 
  Jeff Threat ⛧❤️‍🔥 (@JeffThreat) reported 22 minutes ago

  @AmazonHelp The driver threw my package into water on my driveway, it was raining... the package busted open and my item went floating down the street. Neighbors saw this and retrieved the item, I was at work.

• 
  Karen Porter (@buckeyeklp) reported 24 minutes ago

  @AmazonHelp It was taken care of and money refunded with no issues but thank you for responding. It was just disappointing to not have the item.

• 
  William Gouge (@William_Gouge_X) reported 25 minutes ago

  @BereanHouse @MikeWingerii I consider my self a 99% cessationist. I don’t deny God has the power to perform miracles today… I’ve experienced them myself… but not the office of apostle. I believe God miraculously heals sometimes (when he desires, and sometimes a result of heavy prayer), but the “gift of faith healer” like the apostles doesn’t exist today in the church. I believe it’s possible in the Amazon for some stranded missionary to speak and unknown language if God so willed… but rolling around on the floor barking like dogs and yelling gibberish is not it. That stuff didn’t exist for 1900 years after the apostles… these modern guys(and girls) just want “spiritual power” no different than Simon the sorcerer in Acts… they get off on power and prestige. None of them are healing kids at St Jude’s… all the demon slayers are constantly casting out demons mint their own congregants inside the church….. hahaha. Why does no one ask why there are so many demons in their churches they’re casting them out by the dozens every week. Make it make sense! Are the demons voluntarily forcing their NPCs to go to the demon slayers churches… if we are going to be honest, most of them are fakers… but some of them are demonic themselves. That Vladchuk guy or whatever. Here’s my point… Mike will take down the obvious guys that already slipped up and got caught… but there is so much craziness in that movement… you CAN NOT separate it from the foundational theology and doctrine it grew out of… it’s all connected… but everyone I. That camp just wink and shrug at this crazy stuff in their churches… 300 years ago, the other Christian’s in town would be burning down that church and kicking people out of the country… and Mike is upset over a mean tweet by Joel to some lNdian invader into our country… we need more bold and brash men I. The church… effeminate men like Mike Winger, constantly with their ******* in a *** and offended over a “bullying tweet” can just go sit out the battle for Christianity on the horizon… he’s a serious liability and I don’t trust him to watch my flank without getting stabbed in the back over a mean tweet… my two cents bro. Enjoy your Lord’s Day.

• 
  FullPortCapital (@FullPortCapital) reported 28 minutes ago

  @Edward_Vamanos @homesteadpilot “Working” lol okay, he’s a cheap labor slave for Amazon, he needs to be deported. Another mouth to feed that doesn’t belong. If it’s brown flush it down, adios puta

• 
  Stonk Daddy (@guccigoy69) reported 28 minutes ago

  @TheBobGob123 @ruthdelioncourt @Savsays Amazon is completely overrun by Indians. I used to work there and there was a tool where you can search any Amazon employee and you can see their whole chain of command. I went down a rabbit hole, looking and managers of managers of managers and it gets VERY Indian the higher up

• 
  Ben Hogan (@benhogan79) reported 31 minutes ago

  @zzzLLOYDYzzz Should have got on the Amazon app mate reckon I got the whole setup for about $650 (excluding PC) Computer table for about $150, bagged a good monitor at 50% off for $250 and a typhoon chair for about $250. Got his PC from Techfast in SA for around $1,500 Havent had a problem.

• 
  John Paterson (@JohnPaterson24) reported 38 minutes ago

  Figured I’d throw out some Sunday ramblings to collect a few thoughts I’ve had almost a month into this. I started bidding the first week of July while we were getting set up at the warehouse and had a few delays along the way. We’re approaching the 2 week mark for having liquidation inventory. Man it’s a lot of work, which is fine and what I expected. What I didn’t expect was not being able to turn inventory over faster. Some of that is me needing to improve my listings which I’m working on, but just based on the interactions I’ve had with potential customers I think it’s more about price than anything. Things seem to change drastically customer wise above $1k, I think the crowd that can afford paying more would rather just deal with an actual store honestly. And trying to sell $3k patio furniture at half price that has scratches is challenging. So that leads me to the cliche of make money on the buy, which is 100% true. I ended up around 30% for most of what I have which reflecting was too high. You need to be able to get top of market when you sell and need to pray everything is in reasonably sellable condition. Neither of those has happened which I should have seen coming quite frankly. The whole exercise has me questioning my next move, and not because I don’t think you can make good money in liquidation, my timeline is filled with people killing it which I love to see. But the math changes with a warehouse and paying employees, and it really requires some smart buying. Maybe that is solved by where I source next, but I already see some challenges with moving to a lower dollar but higher volume type model for MY SPECIFIC case. I stress that again especially for the newer people, you guys should be crushing it out of your garage or storage unit with low overhead. I capitalized that part because someone will read this thinking I’m complaining and it’s quite the opposite. This is tuition I expected to pay. But at the same time, I run a successful Amazon business, and part of what eased the risk of the warehouse was the ability to pull back what I send to the prep center and have my son prep it here, which he is good at. The problem is it circles back in part to why I started this, which was to get a good chunk of my eggs out of the Amazon basket. So where that leaves me I’m not quite sure. If I can’t turn and burn higher dollar furniture locally it’s possible it’s not the right model for me, because it does take time from Amazon and I think I could be better and more efficient scaling something like Walmart instead. I wanted faster cash turnover trying local sales because Amazon is incredibly slow, so it’s left me questioning my path. I’m sure some people are reading this rolling their eyes, and I hesitated to share because I get it. Two weeks in and he’s already quitting, I can hear it now lol. I’m really not, but I do need to be intentional with my time. I just had a low Amazon revenue month because my buying was down in June focusing on this new venture. However my ROI was up as we doubled down in a specific area and I had my best buying month in July for inventory. We cranked out a lot of shipments that will be going live in the coming weeks. I’m going to keep sharing regardless because I truly value the connections I’m building, and I’m always big on we can all grow together. Happy selling!

• 
  - (@imsyedshz) reported 39 minutes ago

  @AmazonHelp @amazonIN as i mentioned link is not working!!!!!!!!!!!!!!!!!!

• 
  Amazon Help (@AmazonHelp) reported 40 minutes ago

  @pappu83252698 @amazon @pappu83252698 Hi! We understand your concern here. As informed earlier, for cancelled orders, a refund will be automatically issued to your original payment method within 24 hours from the time the tracking was updated as "undeliverable/returned to seller". Further, it takes an additional 5 business days for the refund to be credited into your original payment method. An email will be sent to you once the refund is successfully processed. More information on refund timelines is shared via the link given above. Kindly keep us posted if the issue isn't resolved in the mentioned time frame, so we can assist you further. We appreciate your patience. Please don’t provide your order/account details as we consider them to be personal information. Our X page is visible to public. -Salwa

• 
  CasualKnowledge (@casualknowledg) reported 42 minutes ago

  Ok so can someone just cooy twitch boot it up and call it twatch and start moving creators overbecause this is getting ridiculous. We need to stop twitch at this point and collectively leave the platform to bankrupt amazon at this point. Or someone buy twitch and get this fix.

• 
  Kris Patel 🇺🇸 (@KrisPatel99) reported 43 minutes ago

  $AMZN Amazon is going to keep spending as long as they see the demand picture remaining strong. I see a lot of accounts post about why $AMZN is cheap based on future operating cash flow projections but I caution people to not ignore counterparty risk. The biggest risk to $AMZN and the wider AI market from chips, to power to neoclouds relies on 2 companies, OpenAI and Anthropic, to be able to generate enough in cashflow to pay their contractual obligations. All of the hyperscalers at one point or another on their conference calls left themselves escape hatches just in case those obligations needed to be modified via slow capex growth and shifting obligations from frontier to a more diverse set of customers. For anyone wondering what I'm talking about, please take a look at Amazon Sagemaker and Amazon Bedrock. A Lot of them spoke about a future where AI models would be smaller and more localized to individual customers based on their own proprietary datasets. This helps $NVDA as the need to training could remain strong beyond just the 2 main frontier labs. Between the possibility "distillation" revenue from China being challenged in the future, ever increasing spend on training runs to stay at the frontier and maintain super premium level pricing, price competition from Open Source and valuation compression from all of the above... I wouldn't ignore the possibility of an implosion post IPO. The Hyperscalers will in my opinion be fine though as they would just elongate the capex cycle and slowly backfill any gaps over time with their existing customer bases. If this were to come to pass, you could see some margin compression to account for discounts that would need to be provided to sell excess compute. If Anthropic and OpenAI end up not being able to fulfill their obligations around the same time, that could mean all of the hyperscalers cutting capex at the same time and thus could crater the AI buildout thesis. For anyone that wants to skip all of the above reading... DONT ignore OpenAI and Anthropic counterparty risk.

• 
  Senthil Velan (@sentvd) reported 43 minutes ago

  @amazonIN @AmazonHelp Please be ashamed to run such a terrible business, don't be too thick skinned in customer satisfaction and service, You guys were the best in customer care a year ago, now you guys have excelled being the worst.

• 
  Carlos DP 🤖🇺🇸 (@carlosdponx) reported 43 minutes ago

  @MarwaEldiwiny They’re not hiding it’s a G1, it’s obviously a G1 and they don’t lie if asked. I don’t advertise explicitly my robots are built on G1 either, because it’s a distraction when marketing like this. The end user doesn’t care. It’s like if you sold a SaaS product and mentioned the servers run on Amazon AWS, it’s irrelevant to the user. Their tweets are marketing posts, not technical reports. Also idk what you mean by the G1 “having a lot of issues”. It’s the most reliable and mature humanoid platform you can actually buy today (and we have no data on reliability of closed platforms).

• 
  Rekter (@Rekter) reported 43 minutes ago

  Overall Market & Sentiment — Macro: The Fed held at 3.50% to 3.75% in a 9 to 3 vote, the fifth consecutive pause, with Hammack, Kashkari, and Logan dissenting for a quarter-point hike. Warsh declined to offer forward guidance, saying he wants markets to move on economic data rather than Fed commentary, and reiterating there is no soft inflation target. Long-end yields did the work instead: the 30-year hit 5.25%, its highest since 2007, and the 10-year topped 4.7%, while the 2-year fell to 4.24%, a steepening that prices inflation risk rather than tightening. June PCE cooled as expected, with headline down 0.1% on the month and 3.7% year over year (from 4.1%), and core up 0.1% monthly and 3.3% annually. Core has now sat at or above 3.3% for four straight months. The consumer is stretching: real spending rose 0.4% against income growth of 0.2%, pushing the saving rate down to 2.7%. September hike odds sit near 57%; the next FOMC is September 15 to 16, with two labor and two inflation reports between. Oil eased as the U.S. and Iran paused strikes, though the Strait of Hormuz remains closed. Fed officials are weighing cutting annual policy meetings below eight, and Warsh signaled the inflation framework could be reviewed next year. Micro: Earnings are extraordinary and increasingly bifurcated. With 304 S&P 500 companies reported, revenue is up 14% and earnings up 47.7% year over year, though that earnings figure is heavily distorted by Alphabet’s one-time $98 billion gain. Every megacap posted double-digit revenue growth, and the market still split them: Microsoft up about 8% to 16% on 43% Azure growth, Amazon up roughly 9% on 37% AWS growth despite raising 2026 capex to $220 billion, Meta down nearly 10% on a raised spend plan, Apple down more than 7% on soft guidance, and Alphabet down 15% earlier on capex fears. Combined 2026 hyperscaler capex now sums to well over $700 billion. Semis are down 23% from highs, memory down 35%, Micron down 41%, and the sector closed its worst month since 2008, even as China’s CXMT surged 531% on debut to become China’s most valuable listed company at $487 billion. 136 more S&P 500 companies report this week. Sentiment: The mood is neither fear nor greed but genuine confusion about who is in charge. A Fed chair who refuses to guide has handed price discovery to the bond market, which responded by pushing the 30-year to a 19-year high while the funds rate stayed frozen. Equity investors, meanwhile, have adopted a new and unforgiving frame on AI: spending is no longer evidence of demand, it is a liability until proven otherwise, and only companies that can point capex at a visible external revenue line get credit. Apple, which barely spends on AI, got graded on the same curve and failed. Crypto sentiment is neutral to soft, with BTC losing its 20-day and ETH holding a trendline. Last week I laid out three Fed scenarios and the actual outcome was a fourth: hold, dissent three ways, and say nothing at all. Warsh’s experiment is now live, and its first result was a 1,153-point Dow drop and a 30-year Treasury back at 2007 levels. If long yields are the transmission channel, every risk asset is now priced off a bond market with no anchor, which is why Wednesday’s move was so violent and Thursday’s reversal on Microsoft equally so. Volatility is the feature, not the bug. Underneath, the real story was the market permanently raising its standard on AI: six megacaps reported, all grew revenue double digits, and the sorting had nothing to do with the size of the spend and everything to do with whether it has a customer attached. Microsoft and Amazon proved it does. Meta could not. Apple got punished for not spending. That is not a rotation, it is a repricing. The first negative July since 2014 arrived alongside the Dow’s fourth straight winning month, as clean a two-speed market as you will get. The Fed handed the keys to the bond market. Watch the 30-year. It’s now the most important price in the world.

• 
  Kentucky Colonel (@TigerTamer8) reported 50 minutes ago

  @brewskidude $12 from Amazon can fix that

• 
  Suriya (@SuriyaXBT) reported 51 minutes ago

  This week in AI and Crypto . A lot happened. - Coldcard firmware bug - hidden since March 2021 - 1367 BTC drained - Total $88M gone - 4585 wallets wiped - updating firmware does not fix it - Leopold Aschenbrenner' Situational Awareness Fund - $20B AI fund - 439% returns through June - KOSPI crashed, margin calls hit - forced to dump $16B in one block trade - Citadel bought everything - same day every stock he sold ripped 13-27% - Apple Q3 2026 - best quarter in company history - $109.4B revenue - stock dropped 9% anyway - reason: memory chip shortage - $373B in market cap erased in one day - Amazon x OpenAI - $50B investment completed - final $21.3B wired after June 30 - no press release - no announcement - found out via SEC filing on a Friday - OpenAI Astra - OpenAI's new unreleased model - 10 unsolved math problems - some open for decades - all solved - total compute cost: $2,000 one of the wildest weeks

• 
  Krylik_ (@cooming_soon0) reported 53 minutes ago

  @AmazonHelp It didn't work; someone please message me so I can share my account details—I speak Turkish. My problem is this: I submitted a request on July 24th, but I haven't received a reply, and my account was closed unfairly.

• 
  🇬🇧 Tom - Investor £130k (@2147mill) reported 54 minutes ago

  Big earnings week continues. $AMD reports Tuesday. $PLTR reports Monday. $AMD: Revenue expected: $11.32 billion. Up 47% year on year. EPS growth of 235% expected. The AI chip challenger to Nvidia. Data centre revenue exploding. $PLTR: Revenue expected: $1.81 billion. Up 80% year on year. Rule of 40 score hit 145% last quarter. Only matched by Nvidia, Micron and SK Hynix. Both inside $SMGB indirectly. Both pure AI plays. Both reporting into a market that’s rewarding revenue growth and punishing capex. Last week Microsoft up 16%. Amazon up 10% and Meta down 10%.

• 
  Dexter K (@kilod22) reported 56 minutes ago

  @southgrace427 Running like Jaime Sommers in slow motion…indeed my six foot Amazon princess

• 
  Noah (@nb_investing) reported 57 minutes ago

  Weekly Portfolio Update: On the week I was down 8.53% but Thursday helped quite a bit to keep me from a horrific week. YTD now I’m at 24.99% marking easily the worst month of my trading career. Lots happened this past week. Rates left unchanged, most of mag 7 reported, and Leopold was sacrificed to citadel. Amazon and Microsoft earnings seemed to confirm that increased capex spend can be worth it for these large companies and that should help nearly everything in my portfolio. SK Hynix and Samsung also reported earnings highlighting the strength of HBM which should help Micron going forward. SNDK earnings are this week and they should blow those out of the water like usual. Other than that I think the rest of this year will go back to rallying high beta names as history predicts. This has been a brutal beat down from my high of 73% in June but I believe we are back on track to finish the year strong.

• 
  the ginger cat from reno 911 (@frivolousisme) reported 58 minutes ago

  @zayzoonNOW I'm not one to fearmonger about illegal immigration and never will be but USAmericans who think it's "bad" right now are in for a reckoning when South Americans have no more growable soil and the amazon has burnt down

• 
  Springblade 🇺🇸 (@steveMmattison) reported 1 hour ago

  @StephenFleming @Walmart I want to see this. We get 2-3 Walmart grocery deliveries every week now, it's incredible! Saves us massive time driving 30 minutes to Walmart then walking around for an hour then driving home. Better than Amazon deliveries, out of milk & chicken breast, no problem, it'll be here at 3pm! We tip $5 per order and still save time & money. And don't get me started on ordering tires and oil from Walmart dot com, the BEST tire selection on the planet shipped right to your front porch. I just toss them in the back seat and head to Walmart Auto on Tuesday morning 8am when nobody else is around. And you get coast-to-coast warranty!

• 
  J🇺🇸 (@GMUW1987) reported 1 hour ago

  @DefiantLs Absolutely guaranteed he used either an iPhone or Android today (Apple-Google). He used Google today. He ordered from Amazon today. He used a Dell PC today (Dell-Microsoft) There has never been the amount of trickle down in the global economy as we see today.