The US Government Has a Microsoft Problem

When Microsoft revealed in January that foreign government hackers had once again breached its systems, the news prompted another round of recriminations about the security posture of the world’s largest tech company.

Despite the angst among policymakers, security experts, and competitors, Microsoft faced no consequences for its latest embarrassing failure. The United States government kept buying and using Microsoft products, and senior officials refused to publicly rebuke the tech giant. It was another reminder of how insulated Microsoft has become from virtually any government accountability, even as the Biden administration vows to make powerful tech firms take more responsibility for America’s cyberdefense.

That state of affairs is unlikely to change even in the wake of a new report by the Cyber Safety Review Board (CSRB), a group of government and industry experts, which lambasts Microsoft for failing to prevent one of the worst hacking incidents in the company’s recent history. The report says Microsoft’s “security culture was inadequate and requires an overhaul.”

Microsoft’s almost untouchable position is the result of several intermingling factors. It is by far the US government’s most important technology supplier, powering computers, document drafting, and email conversations everywhere from the Pentagon to the State Department to the FBI. It is a critical partner in the government’s cyberdefense initiatives, with almost unparalleled insights about hackers’ activities and sweeping capabilities to disrupt their operations. And its executives and lobbyists have relentlessly marketed the company as a leading force for a digitally safer world.

These enviable advantages help explain why senior government officials have refused to criticize Microsoft even as Russian and Chinese government-linked hackers have repeatedly breached the company’s computer systems, according to cybersecurity experts, lawmakers, former government officials, and employees of Microsoft’s competitors.

These people—some of whom requested anonymity to candidly discuss the US government and their industry’s undisputed behemoth—argue that the government’s relationship with Microsoft is crippling Washington’s ability to fend off major cyberattacks that jeopardize sensitive data and threaten vital services. To hear them tell it, Microsoft is overdue for oversight.

A History of Breaches and Controversy

Microsoft has a long track record of security breaches, but the past few years have been particularly bad for the company.

In 2021, Chinese government hackers discovered and used flaws in Microsoft’s email servers to hack the company’s customers, later releasing the flaws publicly to spark a feeding frenzy of attacks. In 2023, China broke into the email accounts of 22 federal agencies, spying on senior State Department officials and Commerce Secretary Gina Raimondo ahead of multiple US delegation trips to Beijing. Three months ago, Microsoft revealed that Russian government hackers had used a simple trick to access the emails of some Microsoft senior executives, cyber experts, and lawyers. Last month, the company said that attack also compromised some of its source code and “secrets” shared between employees and customers. On Thursday, the Cybersecurity and Infrastructure Security Agency (CISA) confirmed that those customers included federal agencies, and issued an emergency directive warning agencies whose emails were exposed to look for signs that the Russian hackers were attempting to use login credentials contained in those emails.

These incidents occurred as security experts were increasingly criticizing Microsoft for failing to promptly and adequately fix flaws in its products. As by far the biggest technology provider for the US government, Microsoft vulnerabilities account for the lion’s share of both newly discovered and most widely used software flaws. Many experts say Microsoft is refusing to make the necessary cybersecurity improvements to keep up with evolving challenges.

Microsoft hasn’t “adapted their level of security investment and their mindset to fit the threat,” says one prominent cyber policy expert. “It’s a huge fuckup by somebody that has the resources and the internal engineering capacity that Microsoft does.”

The Department of Homeland Security’s CSRB endorsed this view in its new report on the 2023 Chinese intrusion, saying Microsoft exhibited “a corporate culture that deprioritized both enterprise security investments and rigorous risk management.” The report also criticized Microsoft for publishing inaccurate information about the possible causes of the latest Chinese intrusion.

The recent breaches reveal Microsoft’s failure to implement basic security defenses, according to multiple experts.

Adam Meyers, senior vice president of intelligence at the security firm CrowdStrike, points to the Russians’ ability to jump from a testing environment to a production environment. “That should never happen,” he says. Another cyber expert who works at a Microsoft competitor highlighted China’s ability to snoop on multiple agencies’ communications through one intrusion, echoing the CSRB report, which criticized Microsoft’s authentication system for allowing broad access with a single sign-in key.

“You don’t hear about these types of breaches coming out of other cloud service providers,” Meyers says.

According to the CSRB report, Microsoft has “not sufficiently prioritized rearchitecting its legacy infrastructure to address the current threat landscape.”

In response to written questions, Microsoft tells WIRED that it’s aggressively improving its security to address recent incidents.

“We are committed to adapting to the evolving threat landscape and partnering across industry and government to defend against these growing and sophisticated global threats,” says Steve Faehl, chief technology officer for Microsoft’s federal security business.

As part of its Secure Future Initiative launched in November, Faehl says, Microsoft has improved its ability to automatically detect and block abuses of employee accounts, begun scanning for more types of sensitive information in network traffic, reduced the access granted by individual authentication keys, and created new authorization requirements for employees seeking to create company accounts.

Microsoft has also redeployed “thousands of engineers” to improve its products and has begun convening senior executives for status updates at least twice weekly, Faehl says.

The new initiative represents Microsoft’s “roadmap and commitments to answer much of what the CSRB report called out as priorities,” Faehl says. Still, Microsoft does not accept that its security culture is broken, as the CSRB report argues. “We very much disagree with this characterization,” Faehl says, “though we do agree that we haven’t been perfect and have work to do.”

A Security Revenue ‘Addiction’

Microsoft has earned special enmity from the cybersecurity community for charging its customers extra for better security protections like threat monitoring, antivirus, and user access management. In January 2023, the company touted that its security division had passed $20 billion in annual revenue.

“Microsoft has shifted to looking at cybersecurity as something that’s meant to generate revenue for them,” says Juan Andrés Guerrero-Saade, associate vice president of research at security firm SentinelOne. His colleague Alex Stamos recently wrote that Microsoft’s “addiction” to this revenue “has seriously warped their product design decisions.”

These tensions exploded into the open in early 2021, as Congress and the new Biden administration scrambled to understand Russia’s far-reaching SolarWinds hacking campaign.

After breaching government networks through SolarWinds software, Moscow’s operatives fooled Microsoft’s cloud platform into granting them expansive access. Because most agencies weren’t paying for Microsoft’s premium service tier, they didn’t have the network activity logs necessary to detect these intrusions. Lawmakers were outraged that Microsoft was charging the government extra for such a basic feature, and Biden administration officials spent the next two and a half years privately urging Microsoft to make log data free for all customers. Microsoft finally agreed to do so last July—eight days after announcing yet another major hack, this one discovered by an agency paying for log data.

Microsoft won’t say if it plans to make other premium security features free for all of its customers. “We continue to raise the built-in security of our products and services to benefit customers,” Faehl says.

Asked about experts’ arguments that Microsoft’s strategy of profiting off of cybersecurity is incompatible with a security-first mindset, Faehl says, “We would disagree with that characterization.”

A System That’s Everywhere

Microsoft’s dominance has prompted concerns that it represents a single point of failure, concentrating America’s technology dependence in such a way that hackers could easily sabotage essential services by targeting one company’s products.

Few services better illustrate the government’s overwhelming dependence on Microsoft—and an area where some experts say a more diversified approach would be safer—than email. A former US cybersecurity official who works at one of Microsoft’s competitors predicts that an attack crippling Microsoft’s email platform would significantly reduce the government’s ability to operate.

Warnings about a Microsoft “monoculture” date back two decades, but the idea is now attracting new attention from policymakers.

“The US government’s dependence on Microsoft poses a serious threat to US national security,” says US senator Ron Wyden. “The government is effectively stuck with the company’s products, despite multiple serious breaches of US government systems by foreign hackers caused by the company’s negligence.”

Last Monday, Wyden announced draft legislation that would set a four-year deadline for the federal government to stop buying collaboration technology like Microsoft Office that critics say doesn’t integrate well with competing services.

Reducing the government’s reliance on a single vendor wouldn’t just benefit the government, experts say. It would also spread the attack risk across more companies, taking some of the pressure off of Microsoft to protect such a vast portfolio of systems. The giant target on Microsoft’s back makes it a magnet for cybercriminals and government hackers, which helps explain its outsize number of breaches.

The government’s reliance on Microsoft also entrenches a sense of familiarity with its products that cements its places in federal networks. While some agencies are exploring alternatives to Microsoft, most of them are sticking with what they know—largely because it’s easier than switching to an alternative platform, the former cyber official says.

Microsoft denies making it difficult for customers to switch to or incorporate competitors’ products. “Our competitors often stoke subjective complaints about ‘compatibility,’” Faehl says, but “we hear this more from the vendors of some third-party products” than from customers trying to use them.

Regardless, experts say, the upshot is clear: The government is dependent on Microsoft, robbing it of the leverage needed to push back on the company’s practices.

Working the Refs

Microsoft isn’t counting on its market dominance alone to defang government oversight. Since its antitrust battles with the government in the 1990s, the company has crafted a sophisticated public policy strategy that combines earnest calls to protect cyberspace with omnipresent participation in government initiatives.

“Microsoft is by far the slickest operation out there in tech when it comes to these issues,” says Andrew Grotto, a former senior White House cyber official who now leads Stanford University’s Program on Geopolitics, Technology, and Governance and consults for some of Microsoft’s competitors. “They learned this lesson 25 years ago and have been applying it ever since.”

Microsoft’s threat intelligence team, which knows more about malicious cyber activity than virtually any other company and most governments, regularly publishes research about cyber threats and collaborates with law enforcement on operations to dismantle hackers’ infrastructure. The company also helps fund groups like the CyberPeace Institute, which advocates for a safer internet and helps defend nongovernment organizations from hackers. And it has positioned itself as a helpful partner to policymakers who want to take on cyber issues but don’t know where to start, sometimes providing lawmakers with draft legislative language.

With its market dominance and political savvy, Microsoft has ensured that officials almost never publicly criticize it, experts say.

“The government’s uncomfortable saying bad things about Microsoft because they’re fully committed to them,” says Mark Montgomery, senior director of the Center on Cyber and Technology Innovation at the Foundation for Defense of Democracies, a think tank.

The Biden administration has spoken grandly about wielding the government’s formidable contracting power to force companies to improve their security. But with Microsoft, that leverage is nonexistent, experts say. “There is no realistic chance that the government will wholesale cancel its contracts with Microsoft,” Paul Rosenzweig, a cyber consultant and former DHS policy official, says in an email.

Microsoft disputes this argument. “The idea that the government is too dependent on Microsoft is at odds with reality,” Faehl says.

The government’s lack of leverage means federal officials never use the kind of blunt language found in the CSRB report when discussing Microsoft, even when they insist on speaking to reporters anonymously. The result is a remarkable display of government deference to Microsoft.

After Chinese hackers broke into government email systems and eluded agencies not paying for Microsoft’s premium security features, a senior official at CISA acknowledged that Microsoft’s business model was “not yielding the sort of security outcomes that we seek,” but they declined to directly rebuke Microsoft, instead sticking to talking points about productive conversations with the company.

In fact, despite Microsoft’s yearslong defiance of CISA’s high-profile push for companies to be “secure by design,” CISA has steadfastly refused to criticize Microsoft’s failures. When Microsoft finally bowed to pressure and made logs free last July, CISA director Jen Easterly said she was “extremely pleased with Microsoft’s decision.”

The former cyber official finds the government’s meekness remarkable. “When their own emails are stolen, they don’t seem to push back on the vendor who is the cause of that.”

The White House’s National Security Council declined to comment for this story. In a statement, Eric Goldstein, CISA’s executive assistant director for cybersecurity, says his agency “has a robust partnership with Microsoft and will continue to collaborate in many areas,” while also continuing to “impress upon all technology companies the urgency of developing products that are secure by design so that consumers can trust the safety and integrity of the technology that they use every day.”

Microsoft’s Faehl says his company is “committed to secure by design and secure by default.”

Forcing Change

The CSRB report on Microsoft’s cloud breach calls for dramatic changes to the company’s security culture. According to many experts, it’s time for the government to find its spine and compel those changes.

“Big, powerful companies in general don’t change their behavior unless they’re incentivized to do so,” Stanford University’s Grotto says.

The CSRB report recommends tough new requirements for cloud providers like Microsoft, including periodic security reviews after they receive federal contracts. Experts say those requirements could shift corporate incentives in favor of better security.

Microsoft seems to realize that its recent breaches have sparked a public relations crisis. “We expect and welcome fair scrutiny,” says Faehl. “As an industry leader we must be accountable for the security of our products and services.”

At the same time, he says, Microsoft “wouldn’t mind seeing some scrutiny” of its competitors who “seek to sow fear, uncertainty and doubt about our position as a way to seek advantage for their own products.”

Taking on Microsoft would also be a way for the Biden administration to live up to the principles in its National Cybersecurity Strategy, which prioritizes shifting the burden of cybersecurity onto large, well-resourced tech vendors. “They make the point … that this balance needs to shift,” Grotto says. “The question now is, ‘Okay, what does administration do with that diagnosis?’”

There are signs that the administration is heeding this advice. During a briefing with reporters on Thursday about the possibility that Russian operatives stole government secrets through their latest Microsoft hack, Goldstein said that CISA and other agencies are “are working closely with Microsoft, in alignment with the recommendations of the Cyber Safety Review Board, to drive further progress in Microsoft’s improvement plans with their broader security culture and enterprise.”

In the meantime, experts say, the status quo allows Microsoft to shirk responsibility for problems that it is uniquely capable of resolving.

“No harm comes from doing nothing, at least not to these companies,” Guerrero-Saade of SentinelOne says. “And that’s what’s going to destroy us.”