Microsoft has been publishing data about the gender, race, and ethnic breakdown of its employees for more than a decade. Since 2019 it’s been publishing a full diversity and inclusion report annually, and at the same time made reporting on diversity a requirement for employee performance reviews.
Now it’s scrapping its diversity report and dropping diversity and inclusion as a companywide core priority for performance reviews, just months after President Donald Trump issued an executive order to try and eradicate workforce diversity, equity, and inclusion (DEI) initiatives.
Game File reported last week that Microsoft will cease publication of its diversity and inclusion reports this year. “We are not doing a traditional report this year as we’ve evolved beyond that to formats that are more dynamic and accessible — stories, videos, and insights that show inclusion in action,” said Frank Shaw, Microsoft’s chief communications officer, in a statement to Notepad. “Our mission and commitment to our culture and values remain unchanged: empowering every person and organization to achieve more.”
Sources tell me that Microsoft also quietly made some big changes to its employee performance reviews last month, known internally as Connect. Microsoft has removed its companywide security and diversity “core priorities” from its performance reviews, meaning employees no longer have to submit exactly what they did to improve security and diversity and what they plan to do in the future.
Microsoft employees always had to answer “What impact did your actions have in contributing to a more diverse and inclusive Microsoft?” and “What impact did your actions have in contributing to a more secure Microsoft?” Both of these questions have been removed, replaced with a simplified form that asks employees to reflect on the results they delivered and how they achieved them, and any recent setbacks and goals for the future.
The performance review changes were announced through a Viva Engage post on Microsoft’s employee news group, instead of through a mass email. Microsoft described its changes internally as a simplification, and announced that “core priorities are now simply called goals, with at least one goal focused on security.”
In HR documentation, the company doesn’t even use the word “diversity” anymore, opting for just “inclusion” instead. “Security, inclusion, and strong people management remain essential to how we deliver impact at Microsoft,” says Microsoft in its HR documentation. “Inclusion is embedded in how you work, interact, and lead, reflecting our growth mindset culture.”
One employee, who supports Microsoft’s DEI initiatives and wishes to remain anonymous, told me that adding the requirement to its performance reviews five years ago seemed “completely insincere and performative” at the time. “The fact that the company (and most of corporate America) just dropped it proves to me that it was always a shallow commitment.” The employee wants “depth and sincerity” in executing DEI policies, which they say Microsoft never achieved.
Other employees I’ve spoken to about the changes aren’t surprised by Microsoft’s walk back. Some point to Elon Musk’s appearance onstage at Microsoft’s Build conference earlier this year as a sign that Microsoft was cozying up to the Trump administration.
Musk’s appearance at Build in May caused plenty of tension internally, at a time when he was heading up DOGE to dismantle government agencies and government-funded organizations. One source told me at the time that the company’s GLEAM group (Global LGBTQIA+ Employees and Allies at Microsoft) were “incensed” by Microsoft CEO Satya Nadella welcoming Musk to Build.
Musk’s appearance was part of a broader push by Nadella to get Musk’s Grok AI model onboarded to Azure in time for Build. Grok 3 was part of the Build announcements, but months later Microsoft had to cautiously onboard Grok 4 after major concerns about its output. One employee told me over the summer that the safety issues around Grok 4 were “very ugly.”
Microsoft pushed ahead with private testing of Grok 4 with potential enterprise customers, but it also quickly rolled out Grok Code Fast 1 to GitHub Copilot. One Microsoft employee said at the time that “this was pushed out with a rushed security review, a coerced and unwilling engineering team, and in full opposition to our supposed company values.”
It’s now going to be a lot more difficult to judge those company values.
Microsoft is still publishing its Inside Inclusion newsletter and “Code of Us” stories that highlight experiences from Microsoft employees with diverse backgrounds, but they’re not the same as having diversity and inclusion as a core priority for employees, the focus of an annual report, or part of disclosures to shareholders.
LinkedIn has been testing an AI personal assistant for Microsoft execs
Microsoft employees have spotted some company executives using a new, unannounced AI personal assistant. Sources tell me that Microsoft CTO Kevin Scott has “Kevin’s Cosio,” a personal AI assistant that reports directly to him in the company’s global directory, alongside other human direct reports.
Cosio is a project from Microsoft’s LinkedIn team that’s designed to be an autonomous, AI-powered digital assistant. You’d think that would simply be Copilot, but Cosio is described internally as a “next-generation digital worker” that is deeply integrated into Microsoft’s enterprise environment. I’m told it’s capable of automating tasks, building workflows, and collaborating with human employees and other AI agents.
Microsoft has been testing Cosio with some executives internally, including Charles Lamanna, head of Microsoft’s business and industry Copilot (BIC) team. Cosio was part of Microsoft’s new Agent 365 initiative, a framework that controls how AI agents access data and work like human employees. Microsoft was planning to roll it out to all employees by the end of October, but that date passed and only execs have been able to test Cosio.
It sounds like Cosio won’t be rolling out more widely anymore. “As Customer Zero, we’re constantly experimenting with emerging technologies — some make it into our roadmap, others don’t,” says Microsoft spokesperson Cynthia Reynaud, in a statement to Notepad. “Cosio was a helpful experiment that is now informing the features we deliver to customers.”
Still, it’s interesting that Microsoft was testing the viability of having digital worker bots that look like real employees.
- Around 500 million PCs are holding off upgrading to Windows 11, says Dell. We already knew that Windows 11 adoption was slower than Windows 10, but last week Dell put a number on the people holding off upgrading: 500 million. Not only are that many machines capable of being upgraded to Windows 11 but haven’t been, but Dell says around the same amount can’t upgrade due to Windows 11’s strict hardware requirements. I expected millions of consumers to stick with Windows 10, but I didn’t think the overall number (including corporate machines) would be as high as 500 million.
- Microsoft’s latest Windows 11 update improves and breaks dark mode. Microsoft shipped a Windows 11 update this week that was supposed to improve dark mode consistency, but it also added an embarrassing bug. After installing the update, every time you open File Explorer it now flashes white. You couldn’t make this up, and I’m surprised that Microsoft didn’t spot this obvious bug during testing. Microsoft says it’s working on a fix.
- Microsoft’s ugly sweaters return with Clippy, Xbox, and Zune brown options. Microsoft is bringing back its ugly sweaters for the holiday season. This year the company has an “Artifact” sweater with lots of retro iconography, a Zune brown sweater, and even an Xbox option. The Artifact sweater puts Clippy at the center, surrounded by MSN, Minesweeper, Internet Explorer, MS-DOS, and plenty of Windows logos. Because Microsoft can’t resist putting a Copilot logo everywhere, this retro sweater even has a Copilot icon on the sleeve. All the sweaters are available in limited quantities at Microsoft’s online store.
- Satya Nadella warns of AI’s impact on data center power consumption. Microsoft’s CEO admitted this week that the energy use of AI data centers could turn people against the tech industry. In an interview with Mathias Döpfner, the CEO of Axel Springer, Nadella says that the tech industry “needs to earn the social permission to consume energy, because we’re doing good in the world.” He also said that people will accept the extra pressure on the electric grid if it “results in economic growth that is broad-spread in the economy.” So far, we’re still waiting to see if AI is a bubble that will burst or add real value to productivity.
- Microsoft says it’s not lowering sales quotas for AI products. Microsoft has come out swinging against a report from The Information that claimed multiple Microsoft divisions have lowered the sales targets of salespeople for certain AI products after “many of them” missed sales-growth goals. An unnamed Microsoft spokesperson told CNBC that the company has not lowered sales quotas or targets for its salespeople. “Aggregate sales quotas for AI products have not been lowered, as we informed them prior to publication,” said the unnamed Microsoft spokesperson.
- Xbox Cloud Gaming is getting a new design soon. Microsoft is getting ready to revamp the Xbox Cloud Gaming interface with a design that more closely resembles the Xbox PC app. The developer behind Better xCloud spotted the changes, with a promotional video offering a brief look at what’s coming. Given there’s a “try new experience” button in test versions of xCloud, I’d expect we’ll see this appear for Xbox Insiders very soon.
- Linux founder defends Microsoft’s Blue Screen of Death. The creator and lead developer of the Linux kernel, Linus Torvalds, has come out as an unlikely ally to Windows this week. Torvalds defended the Blue Screen of Death errors in Windows in an appearance on Linus Tech Tips, saying that most were related to “hardware being not reliable” instead of software bugs in Windows. Microsoft has now changed the BSOD to black, in an effort to simplify the error screen and to probably shift away from the memes and jokes.
- Microsoft looks to move Xbox production to Vietnam. Microsoft is reportedly planning to move some of its Xbox manufacturing to factories in Vietnam. Reuters reports that a Foxconn subsidiary is seeking a permit to make up to 4.8 million Xbox gaming devices in Vietnamese factories. This follows a report last month that suggested Microsoft was moving its Surface manufacturing out of China and looking to produce more Xbox consoles outside of the country. Moving manufacturing to Vietnam will help Microsoft avoid some of the larger Trump tariffs that have impacted Xbox console prices in the US.
- Microsoft might be ditching Contoso and Fabrikam. I remember doing Microsoft Certified Systems Engineer certifications as a teenager and seeing the fake Contoso and Fabrikam companies in every scenario. Microsoft has used these two companies for demos and testing for decades now, but as the company pushes ahead with its AI transformation it’s also introducing a new fake company: Zava. The Register spotted Microsoft using this company during its Ignite conference, and it has also shown up in some technical documents recently. Microsoft says Zava is a “frontier” company, the moniker it uses for companies that are rapidly adopting AI. Hopefully Contoso and Fabrikam can survive Microsoft’s AI transition, though.
I’m always keen to hear from readers, so please drop a comment here, or you can reach me at [email protected] if you want to discuss anything else. If you’ve heard about any of Microsoft’s secret projects, you can reach me via email at [email protected] or speak to me confidentially on the Signal messaging app, where I’m tomwarren.01. I’m also tomwarren on Telegram, if you’d prefer to chat there.
Thanks for subscribing to Notepad.
