Office 365 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.

Office 365 Issues Reports

Latest outage, problems and issue reports in social media:

• 
  Marco Aurélio (@marcrygamerbr) reported 2 minutes ago

  @Marcelo_gsf_ @davepl1968 @ObnoxiousFumes2 I have invented office365 before microsoft creating a xen server with word, powerpoint, visio and excel hahahaaha, my only windows install

• 
  Jonathan Moore (@jondotgov) reported 2 minutes ago

  As a long time Google Workspace user, I am disappointed at how @Office365 is handling their webapps. I've been struggling with formatting a Word doc all morning, where printing to PDF totally changes the layout... Open in the desktop app and can fix it in 5 min

• 
  Deepraj Samudra (@deepsamudra) reported 9 minutes ago

  @Office365 what's wrong with your support? Registered case#35772656 on 30th March, followed up with support line but no resource to assist on the issue yet.

• 
  Dan Greller (@dgreller) reported 10 minutes ago

  @bretgreenstein @KarolCodes Agreed. Historically, things moved at a slow enough pace that there was ample time to reasonably estimate the cost of an associate's tech stack. The value obtained from that stack was self-evident, exceeding its cost. For example, no rational firm was doing an ROI on whether a new associate should have a PC or an Office 365 license. It was table stakes. In this new AI world, both the costs and the potential value are radically changing on a routine basis. It makes it that much harder to have sound and stable financial governance models.

• 
  Dan Victor, CFA (@DanV_Barrons) reported 11 minutes ago

  @buccocapital exact same scenario playing out with MSFT down 32% from peak. break out the consumer and enterprise facing"business productivity (office 365) facing same AI disruption multiple compression

• 
  Noah (@aotearoanoah) reported 12 minutes ago

  @JamesHalcrow I mean yeah valid. I had issues in HS with the school Office365, the whole schools services got hacked at one stage too and a bunch of seniors lost their work, so I’ve been using personal since

• 
  Fred Says (@FreddSays) reported 13 minutes ago

  Looks like you may be using Office 365 which is online. Better to buy a one-time purchase to install on your PC or Mac. Office Home 2024: £119.99 Office Home & Business: £249.99 Fully supported with updates and trouble shooting.

• 
  Shane (@Shane_L) reported 13 minutes ago

  @adahstwt Just deployed a small app for one of my company’s franchisees to use. Deployed on Azure because we use Office 365 and I could rely on existing MS accounts for auth. Never again… terrible experience. Azure is so disjointed…

• 
  Magical Joestar (@n1qh13) reported 13 minutes ago

  @S4m4_VT @Pirat_Nation Not a Linux problem, 99% of games do works on Linux tho but via workaround (That got simplified by steam in a remarkable way) As for softwares, welp you're not going to use Office 365. You have to adapt and find the right tools that works with your machine.

• 
  puffybsd (@puffybsd) reported 14 minutes ago

  Loading a 2.6mb excel file: * Default chrome plugin on chromebook: unbearably slow. * Office365: faster, but weird placeholder icons. * sheets: fast. (Heard that recent changes will speed office up and remove the progress loading icons.)

• 
  Simon Holman (@SimonHolman) reported 19 minutes ago

  @Outlook I know you want everyone paying for Office365, but do you really need to make configuring IMAP accounts so difficult. Just give the option to specify the servers without just assuming them. I shouldn't have to go into Control Panel to fix it!

• 
  Florian Lohoff (@fl0h0ff) reported 25 minutes ago

  Hey @Office365 - How do i fix broken and completely outdated records in the acompli/outlookmobile.com autodiscover database? Customers are getting redirected to gmail which is not the tennant anymore.

• 
  Looks Safe To Me (@Mackama_Too) reported 27 minutes ago

  @codewith55 Your first step for a new Windows computer is not what to install, but what to uninstall. 1) OneDrive 2) MS-Account (use local login) 3) Copilot 4) Office365 5) Manufacturer & MS Bloatware. Then start your installs. 1) A real browser (Brave, Firefox) 2) LibreOffice 3) VLC

• 
  DataJuggler (@DataJuggler007) reported 31 minutes ago

  @FurkanGozukara I stopped giving money to companies I can't call. In July 2024, Microsoft expected me to go a month without email. I don't own a cell phone, email is my communication with the world. I cancelled Office 365, moved my mx records for my domain and went down to $1.50 a month.

• 
  💯 Almin Ibrahimović MBA CGA FCMI CMgr CITP MBCS (@_almin) reported 31 minutes ago

  @Microsoft365 To use Scout, you need: 1. to be part of the Frontier Preview Program 2. need Office 365 Copilot Access 3. need local admin privileges on the device 4. your IT Admin needs to configure Frontier. 5. your IT Admin needs to configure InTune policy for Scout 6. your local IT Admin needs to agree to (attest to) and opt-in their organisation to third party inference paths. Note: "When data is sent to GitHub Copilot, M365 data residency, retention, eDiscovery, legal hold, Data Loss Prevention (DLP), Sensitivity and Confidentiality Labels, audit, SLAs, and the Customer Copyright Commitment do not apply." Here is why this matters: If 'AI' extracts a list of sensitive intellectual property or customer data from a labeled document and pastes it into a new response, that new response has no label. A user could easily copy and paste that text into an external email or a public generative AI tool without triggering any data loss prevention alerts. Regulatory frameworks require strict auditing of where sensitive data travels. Because the new content does not inherit the source classification, your automated compliance systems lose track of that data. The responsibility shifts entirely to the end user to manually classify every single output generated by the AI session, which frequently leads to human error.