A scandal unfolding in Argentina shows the dangers of implementing facial recognition—even with laws and limits in place.

The Twisted Eye in the Sky Over Buenos Aires

Signs suggest the culprits worked within a notorious Chinese hacker group that may have also hacked Indian electric utilities years earlier.

The loose nexus of Chinese-origin cyberspies collectively called APT41 is known for carrying out some of the most brazen hacking schemes linked to China over the past decade. Its methods range from a spree of software supply chain attacks that planted malware in popular applications to a sideline in profit-focused cybercrime that went so far as to steal pandemic relief funds from the US government. Now, an apparent offshoot of the group appears to have turned its focus to another worrying category of target: power grids.

Today, researchers on the Threat Hunter Team at Broadcom-owned security firm Symantec revealed that a Chinese hacker group with connections to APT41, which Symantec is calling RedFly, breached the computer network of a national power grid in an Asian country—though Symantec has declined to name which country was targeted. The breach began in February of this year and persisted for at least six months as the hackers expanded their foothold throughout the IT network of the country's national electric utility, though it's not clear how close the hackers came to gaining the ability to disrupt power generation or transmission.

The unnamed country whose grid was targeted in the breach was one that China would “have an interest in from a strategic perspective,” hints Dick O'Brien, a principal intelligence analyst on Symantec's research team. O'Brien notes that Symantec doesn't have direct evidence that the hackers were focused on sabotaging the country's grid, and says it's possible they were merely carrying out espionage. But other researchers at security firm Mandiant point to clues that these hackers may be the same ones that had been previously discovered targeting electrical utilities in India. And given recent warnings about China's hackers breaching power grid networks in US states and in Guam—and specifically laying the groundwork to cause blackouts there—O'Brien warns there's reason to believe China may be doing the same in this case.

“There are all sorts of reasons for attacking critical national infrastructure targets,” says O'Brien. “But you always have to wonder if one \[reason\] is to be able to retain a disruptive capability. I'm not saying they would use it. But if there are tensions between the two countries, you can push the button.”

Symantec's discovery comes on the heels of warnings from Microsoft and US agencies including the Cybersecurity and Infrastructure Security Agency (CISA) and the National Security Agency (NSA) that a different Chinese state-sponsored hacking group known as Volt Typhoon had penetrated US electric utilities, including in the US territory of Guam—perhaps laying the groundwork for cyberattacks in the event of a conflict, such as a military confrontation over Taiwan.  _The New York Times_ later reported that government officials were particularly concerned that the malware had been placed in those networks to create the ability to cut power to US military bases.

In fact, fears of a renewed Chinese interest in hacking power grids stretch back to two years ago, when cybersecurity firm Recorded Future warned in February 2021 that Chinese state-sponsored hackers had placed malware in power grid networks in neighboring India—as well as railways and seaport networks—in the midst of a border dispute between the two countries. Recorded Future wrote at the time that the breach appeared to be aimed at gaining the ability to cause blackouts in India, though the firm said it wasn't clear whether the tactic was designed to send a message to India or to gain a practical capability in advance of military conflict, or both.

Some evidence suggests the 2021 India-focused hacking campaign and the new power grid breach identified by Symantec were both carried out by the same team of hackers with links to the broad umbrella group of Chinese state-sponsored spies known as APT41, which is sometimes called Wicked Panda or Barium. Symantec notes that the hackers whose grid-hacking intrusion it tracked used a piece of malware known as ShadowPad, which was deployed by an APT41 subgroup in 2017 to infect machines in a supply chain attack that corrupted code distributed by networking software firm NetSarang and in several incidents since then. In 2020, five alleged members of APT41 were indicted and identified as working for a contractor for China's Ministry of State Security known as Chengdu 404. But even just last year, the US Secret Service warned that hackers within APT41 had stolen millions in US Covid-19 relief funds, a rare instance of state-sponsored cybercrime targeting another government.

Although Symantec didn't link the grid-hacking group it's calling RedFly to any specific subgroup of APT41, researchers at cybersecurity firm Mandiant point out that both the RedFly breach and the years-earlier Indian grid-hacking campaign used the same domain as a command-and-control server for their malware: Websencl.com. That suggests the RedFly group may in fact be tied to both cases of grid hacking, says John Hultquist, who leads threat intelligence at Mandiant. (Given that Symantec wouldn't name the Asian country whose grid RedFly targeted, Hultquist adds that it may in fact be India again.)

More broadly, Hultquist sees the RedFly breach as a troubling sign that China is shifting its focus toward more aggressive targeting of critical infrastructure like power grids. For years, China largely focused its state-sponsored hacking on espionage, even as other nations like Russia and Iran have attempted to breach electrical utilities in apparent attempts to plant malware capable of triggering tactical blackouts. The Russian military intelligence group Sandworm, for example, has attempted to cause three blackouts in Ukraine—two of which succeeded. Another Russian group tied to its FSB intelligence agency known as Berserk Bear has repeatedly breached the US power grid to gain a similar capability, but without ever attempting to cause a disruption.

Given this most recent Chinese grid breach, Hultquist argues it's now beginning to appear that some Chinese hacker teams may have a similar mission to that Berserk Bear group: to maintain access, plant the malware necessary for sabotage, and wait for the order to deliver the payload of that cyberattack at a strategic moment. And that mission means the hackers Symantec caught inside the unnamed Asian country's grid will almost certainly return, he says.

“They have to maintain access, which means they're probably going to go right back in there. They get caught, they retool, and they show up again,” says Hultquist. “The major factor here is their ability to just stay on target—until it's time to pull the trigger.”

China-Linked Hackers Breached a Power Grid—Again

State and local governments in the US are scrambling to harness tools like ChatGPT to unburden their bureaucracies, rushing to write their own rules—and avoid generative AI's many pitfalls.

The United States Environmental Protection Agency blocked its employees from accessing ChatGPT while the US State Department staff in Guinea used it to draft speeches and social media posts.

Maine banned its executive branch employees from using generative artificial intelligence for the rest of the year out of concern for the state’s cybersecurity. In nearby Vermont, government workers are using it to learn new programming languages and write internal-facing code, according to Josiah Raiche, the state’s director of artificial intelligence.

The city of San Jose, California, wrote 23 pages of guidelines on generative AI and requires municipal employees to fill out a form every time they use a tool like ChatGPT, Bard, or Midjourney. Less than an hour’s drive north, Alameda County’s government has held sessions to educate employees about generative AI’s risks—such as its propensity for spitting out convincing but inaccurate information—but doesn’t see the need yet for a formal policy.

“We’re more about what you can do, not what you can’t do,” says Sybil Gurney, Alameda County’s assistant chief information officer. County staff are “doing a lot of their written work using ChatGPT,” Gurney adds, and have used Salesforce’s Einstein GPT to simulate users for IT system tests.

At every level, governments are searching for ways to harness generative AI. State and city officials told WIRED they believe the technology can improve some of bureaucracy’s most annoying qualities by streamlining routine paperwork and improving the public’s ability to access and understand dense government material. But governments—subject to strict transparency laws, elections, and a sense of civic responsibility—also face a set of challenges distinct from the private sector.

“Everybody cares about accountability, but it’s ramped up to a different level when you are literally the government,” says Jim Loter, interim chief technology officer for the city of Seattle, which released preliminary generative AI guidelines for its employees in April. “The decisions that government makes can affect people in pretty profound ways and … we owe it to our public to be equitable and responsible in the actions we take and open about the methods that inform decisions.”

The stakes for government employees were illustrated last month when an assistant superintendent in Mason City, Iowa, was thrown into the national spotlight for using ChatGPT as an initial step in determining which books should be removed from the district’s libraries because they contained descriptions of sex acts. The book removals were required under a recently enacted state law.

That level of scrutiny of government officials is likely to continue. In their generative AI policies, the cities of San Jose and Seattle and the state of Washington have all warned staff that any information entered as a prompt into a generative AI tool automatically becomes subject to disclosure under public record laws.

That information also automatically gets ingested into the corporate databases used to train generative AI tools and can potentially get spit back out to another person using a model trained on the same data set. In fact, a large Stanford Institute for Human-Centered Artificial Intelligence study published last November suggests that the more accurate large language models are, the more prone they are to regurgitate whole blocks of content from their training sets.

That’s a particular challenge for health care and criminal justice agencies.

Loter says Seattle employees have considered using generative AI to summarize lengthy investigative reports from the city’s Office of Police Accountability. Those reports can contain information that’s public but still sensitive.

Staff at the Maricopa County Superior Court in Arizona use generative AI tools to write internal code and generate document templates. They haven’t yet used it for public-facing communications but believe it has potential to make legal documents more readable for non-lawyers, says Aaron Judy, the court’s chief of innovation and AI. Staff could theoretically input public information about a court case into a generative AI tool to create a press release without violating any court policies, but, she says, “they would probably be nervous.”

“You are using citizen input to train a private entity’s money engine so that they can make more money,” Judy says. “I’m not saying that’s a bad thing, but we all have to be comfortable at the end of the day saying, ‘Yeah, that’s what we’re doing.’”

Under San Jose’s guidelines, using generative AI to create a document for public consumption isn’t outright prohibited, but it is considered “high risk” due to the technology’s potential for introducing misinformation and because the city is precise about the way it communicates. For example, a large language model asked to write a press release might use the word “citizens” to describe people living in San Jose, but the city uses only the word “residents” in its communications. because not everyone in the city is a US citizen.

Civic technology companies like Zencity have added generative AI tools for writing government press releases to their product lines, while the tech giants and major consultancies—including Microsoft, Google, Deloitte, and Accenture—are pitching a variety of generative AI products at the federal level.

The earliest government policies on generative AI have come from cities and states, and the authors of several of those policies told WIRED they’re eager to learn from other agencies and improve their standards. Alexandra Reeve Givens, president and CEO of the Center for Democracy and Technology, says the situation is ripe for “clear leadership” and “specific, detailed guidance from the federal government.”

The federal Office of Management and Budget is due to release its draft guidance for the federal government’s use of AI some time this summer.

The first wave of generative AI policies released by city and state agencies are interim measures that officials say will be evaluated over the coming months and expanded upon. They all prohibit employees from using sensitive and non-public information in prompts and require some level of human fact checking and review of AI-generated work, but there are also notable differences.

For example, guidelines in San Jose, Seattle, Boston, and the state of Washington require that employees disclose their use of generative AI in their work product while Kansas’ guidelines do not.

Albert Gehami, San Jose’s privacy officer, says the rules in his city and others will evolve significantly in coming months as the use cases become clearer and public servants discover the ways generative AI is different from already ubiquitous technologies.

“When you work with Google, you type something in and you get a wall of different viewpoints, and we’ve had 20 years of just trial by fire basically to learn how to use that responsibility, “ Gehami says. “Twenty years down the line, we’ll probably have figured it out with generative AI, but I don’t want us to fumble the city for 20 years to figure that out.”

AI Chatbots Are Invading Your Local Government—and Making Everyone Nervous

Plus: Apple patches newly discovered flaws exploited by NSO Group spyware, North Korean hackers target security researchers, and more.

Last week, WIRED published a deep-dive investigation into Trickbot, the prolific Russian ransomware gang. This week, US and UK authorities sanctioned 11 alleged members of Trickbot and its related group, Conti, including Maksim Galochkin, aka Bentley, one of the alleged members whose real-world identity we confirmed through our investigation. Coincidence? Maybe. Either way, it's a big deal.

In addition to the US and UK sanctions, the US Justice Department also unsealed indictments filed in three US federal courts against Galochkin and eight other alleged Trickbot members for ransomware attacks against entities in Ohio, Tennessee, and California. Because everyone charged is a Russian national, however, it is unlikely they will ever be arrested or face trial.

While Russian cybercriminals typically enjoy immunity, the same may not remain true for the country’s military hackers. The lead prosecutor of the International Criminal Court (ICC) says the ICC will begin pursuing charges for cyber war crimes. The prosecutor, Karim Khan, did not name Russia, but the move follows a formal petition from the Human Rights Center at UC Berkeley’s School of Law asking the ICC to prosecute Russia’s Sandworm hackers for war crimes. Part of Russia’s GRU military intelligence agency, Sandworm is responsible for causing blackouts in Ukraine, the only known instances of cyberattacks shutting down an electrical grid. Sandworm also released the NotPetya malware against Ukraine, which ultimately spread globally and caused an unprecedented $10 billion in damages worldwide.

Russia is far from the only country that engages in offensive cyberwar tactics. China-backed hackers have repeatedly targeted the US and other countries, and they may be getting some help finding unpatched vulnerabilities. A Chinese law passed in 2022 demands that any network technology company operating in the country share details about vulnerabilities in its products with the Chinese government within two days of their discovery. Information about these vulnerabilities may then be shared with China’s hackers. It’s unclear how many Western companies comply with the law or provide enough information to allow Chinese hackers to exploit the products’ flaws.

Speaking of Chinese hackers, Microsoft this week finally explained how China’s state-sponsored hackers managed to steal a cryptographic key that allowed the attackers to successfully access the Outlook email accounts of at least 25 organizations, including US government agencies. According to Microsoft, the hackers broke into the account of a company engineer using token-stealing malware. They then used that account to access a cache of crash data that accidentally contained the signing key they then stole and used to go on an Outlook hacking spree. None of this was supposed to be possible, and Microsoft says it has corrected several flaws in its systems that allowed the attack to happen.

Before he died in a mysterious plane crash last month following an attempted coup against Russian president Vladimir Putin, Yevgeny Prigozhin wasn’t just the leader of the Wagner Group mercenaries. He was also the head of the notorious Internet Research Agency (IRA), a Russian outfit responsible for widespread disinformation campaigns. While the IRA was reportedly shut down, new research shows that pro-Prigozhin trolls continue to push his agenda. Many of the accounts spreading disinformation on X (formerly Twitter) have been banned. But since when has that stopped them?

Elsewhere, we explained how prompt injection attacks against generative AI chatbots like ChatGPT take advantage of a flaw that’s difficult to fix. We detailed how hard it is to opt out of allowing Facebook to use your data to train its AI. We have a rundown on Proton Sentinel, a suite of tools that are similar to Google’s offerings but with a strong emphasis on privacy and security. We also co-published a story with The Markup into Axon’s quest to build Taser-armed drones. And we got the inside scoop on a meeting between top US spies and civil liberties groups over Section 702 of the Foreign Surveillance Intelligence Act, which is set to expire at the end of the year.

But that’s not all. Each week, we round up the security and privacy news we didn’t cover in depth ourselves. Click the headlines to read the full stories. And stay safe out there.

Car companies are collecting and selling extremely detailed personal data from drivers who have no real way to opt out, a new report from the Mozilla Foundation found. Researchers spent hundreds of hours studying 25 privacy policies for major car brands and found that none of them met the foundation’s minimum standards around privacy and security.

According to the report, modern cars, stuffed to the roof with sensors, collect more information about you than just about any other product in your life. They know where you go, what you say, and how you move your body. Nissan’s privacy policy, for example, allows the company to collect and share drivers’ sexual activity, health diagnosis data, and genetic information, according to the report.

Eighty-four percent of the brands that researchers studied share or sell this kind of personal data, and only two of them allow drivers to have their data deleted. While it is unclear exactly who these companies share or sell data to, the report points out that there is a huge market for driver data. An automotive data broker called High Mobility cited in the report has a partnership with nine of the car brands Mozilla studied. On its website, it advertises a wide range of data products—including precise location data.

This isn’t just a privacy nightmare but a security one. Volkswagen, Toyota, and Mercedes-Benz have all recently suffered data leaks or breaches that affected millions of customers. According to Mozilla, cars are the worst category of products for privacy that they have ever reviewed.

Apple has just released a security update to iOS after researchers at Citizen Lab discovered a zero-click vulnerability being used to deliver Pegasus spyware. Citizen Lab, which is part of the University of Toronto, is calling the newly discovered exploit chain Blastpass. Researchers say it is capable of compromising iPhones running the latest version of iOS (16.6) without the target even touching their device. According to researchers, Blastpass is delivered to a victim’s phone through an iMessage with an Apple Wallet attachment containing a malicious image.

The Pegasus spyware, developed by NSO Group, enables an attacker to read a target’s text messages, view their photos, and listen to calls. It has been used to track journalists, political dissidents, and human rights activists around the world.

Apple says customers should update their phones to the newly released iOS 16.6.1. The exploit can also attack certain models of iPads. You can see details of the affected models here. Citizen Lab urges at-risk users to enable Lockdown Mode.

North Korea-backed hackers are targeting cybersecurity researchers in a new campaign that is exploiting at least one zero-day vulnerability, Google’s Threat Analysis Group (TAG) warned in a report released Thursday. The group did not provide details about the vulnerability since it is currently unpatched. However, the company says it is part of a popular software package used by security researchers.

According to TAG, the current attack mirrors a January 2021 campaign that similarly targeted security researchers working on vulnerability research and development. Like the previous campaign, North Korean threat actors send researchers malicious files after first spending weeks establishing a relationship with their target. According to the report, the malicious file will execute “a series of anti-virtual machine checks” and send collected information—along with a screenshot—back to the attacker.

In order to shield prospective jurors from harassment, District Attorney Fani Willis asked the judge in Donald Trump’s racketeering trial to prevent people from capturing or distributing any sort of image or identifying information about them. The motion, filed in Fulton County Superior Court on Wednesday, revealed that immediately after the indictment was filed, anonymous individuals on “conspiracy theory websites" had shared the full names, ages, and addresses of 23 grand jurors with “the intent to harass and intimidate them.”

Willis also revealed that she had been the victim of doxxing when the personal information of her and her family—including their physical addresses and “GPS coordinates”—was posted on an unnamed website hosted by a Russian company. Willis, who is Black, had previously disclosed that she faced racist and violent threats after the announcement of her investigation into the former president.

Mozilla: Your New Car Is a Data Privacy Nightmare

Civil rights groups say efforts to get US intelligence agencies to adopt privacy reforms have largely failed. Without those changes, renewal of a post-911 surveillance policy may be doomed.

Senior United States intelligence officials met privately in Virginia yesterday with over a dozen civil liberties groups to field concerns about domestic surveillance operations that have drawn intense scrutiny this summer among an unlikely coalition of Democratic and Republican lawmakers in the US Congress.

The closed-door session, convened at the Liberty Crossing Intelligence Campus—a sprawling complex housing the bulk of the nation’s counterterrorism infrastructure—comes amid a backdrop of political furor over past misuses of a powerful surveillance tool by, principally, the Federal Bureau of Investigation (FBI). Republican lawmakers, who remain aggrieved over the FBI’s botched operation to surveil a former Trump campaign aide amid its 2016 Russia investigation, have formed an extraordinary alliance with Democratic rivals who’ve long been critical of the FBI’s power to warrantlessly access information about Americans “incidentally” collected by spies in the process of monitoring foreign threats.

The meeting, organized by the director of national intelligence, Avril Haines, was attended by top officials from the National Security Agency (NSA), US Department of Justice (DOJ), and Central Intelligence Agency (CIA), among others. General Paul Nakasone, the NSA director, is believed to have attended, though neither the IC, nor any source at the meeting, would confirm or deny his presence. (All sources spoke with WIRED on background citing rules established ahead of the gathering.)

Privacy and civil liberties advocates in attendance Thursday say one of their chief objectives was putting the intelligence community (IC) on notice: Without significant privacy reforms, any effort to reauthorize the use of its most powerful surveillance weapon—Section 702 of the Foreign Intelligence Surveillance Act—will be a doomed undertaking. The 9/11-era program, occasionally referred to as the “crown jewel” of US intelligence, is set to expire at the end of the year. Sources in Congress with knowledge of ongoing negotiations over the program say Biden administration officials have privately encouraged lawmakers to pass a “clean bill” this winter, airing fears that any potential lapse in surveillance would pose a national security threat. Targets of the 702 program have expanded in the past decade beyond terrorists in the Middle East, and today include foreign cybersecurity threats linked to Iran, Russia, and China, as well as drug traffickers involved in the production of fentanyl, a dangerous opioid flooding US streets.

The fate of the 702 program hangs by a precarious thread, with lawmakers on both sides of the aisle increasingly scrutinous of the FBI’s ability to tap into data that the intelligence community has long claimed is only unintentionally collected on Americans—a byproduct of casting a wide surveillance net over the communications of hundreds of thousands of individuals each year believed or assumed to be agents of hostile foreign powers. Restricting the bureau’s access to this data for domestic criminal investigation without first obtaining a court order remains one of the top reforms sought after by IC’s bipartisan critics.

Sources at the meeting say the conversation was largely one-sided, with Haines and other intelligence officials framing the event as purely an opportunity to bear witness to the concerns of civil rights advocates. While none expected a true back-and-forth discussion, some advocates nevertheless expressed frustration over the lack of reciprocity, with one describing it bluntly as “stonewalling.” A spokesperson for the IC said such “listening sessions,” in which top officials gather to bear witness to the concerns of relevant civil society stakeholders, are commonplace, and that, generally speaking, the IC does not disclose the nature of its conversations with members of Congress.

Its guests on Thursday included privacy and national security experts from the American Civil Liberties Union, Brennan Center for Justice at NYU School of Law, Electronic Information Privacy Center, and Demand Progress, among a dozen other groups. The largely progressive coalition further included conservative nonprofits such as FreedomWorks and Americans for Prosperity. Bob Goodlatte, a former Republican chair of the House Judiciary Committee who now serves as a senior advisor to the nonprofit Project for Privacy and Surveillance Accountability, also attended.

While Haines and some of the officials under her have expressed a willingness to enhance safeguards around intelligence “incidentally” collected on Americans, the IC has been unwilling so far to disclose what, if any, specific reforms it may be receptive to adopting. Privacy advocates say they remain, in response, skeptical it’s truly open to reform, adding that key sources on Capitol Hill continue to be unaware of any specific reforms currently supported by IC. In a letter ahead of the meeting, the same advocates said they were pursuing new limits on the scope of US surveillance codified under federal law, and the closing of a notorious loophole enabling intelligence and law enforcement services to bypass warrant requirements by paying data brokers for Americans’ sensitive data.

An internal advisory report declassified by Haines earlier this year described “large” purchases of “sensitive and intimate information” about Americans from private companies, including data enabling law enforcement to trace Americans’ whereabouts over extended periods of time. The report acknowledges that much of the purchased data would ordinarily fall under the protection of the US Constitution's Fourth Amendment, which includes a guarantee against unreasonable searches and seizures. Various intelligence agencies, the Defense Intelligence Agency and NSA among them, have taken the position that paying for access to sensitive data, as opposed to demanding it in accordance with criminal law procedures, effectively negates that constitutional right.

Defenders of this particular collection method argue that the data is already widely available to private companies and even foreign intelligence services. Its detractors note that companies and foreign countries are not bound by the Fourth Amendment, and otherwise lack the US government’s power to arrest, fine, and imprison US residents.

Sources say Haines and other officials were pressed to disclose whether the government has arrived at a single, coherent policy regarding commercial data purchases. It is currently assumed US spy agencies are left to determine the legality of such arrangements individually, with declassified materials over the past year overwhelmingly supporting that belief. None of the IC officials, sources say, would comment on whether such a policy exists, though Haines has previously told members of Congress that devising one is a process that’s been underway since early 2021.

A spokesperson at the Office of the Director of National Intelligence (ODNI) says Haines had a prior speaking engagement Friday morning and could not be immediately reached for comment.

In a statement, members of the privacy coalition expressed gratitude for Haines’ taking the time to hear their concerns. Nevertheless, civil liberties experts say they remain “deeply distressed” by the IC’s unwillingness to commit to any “meaningful reforms that are critical to protect Americans’ privacy,” calling specifically the FBI’s abuses of 702-related data. “The administration and intelligence community must be willing to come to the table and accept significant new privacy protections that advocates, Congress, and the American people are calling for,” they said.

Top US Spies Meet With Privacy Experts Over Surveillance 'Crown Jewel'

The CEO’s vision for Taser-equipped drones includes a fictitious scenario in which the technology averts a shooting at a day care center.

_This article was_ _copublished_ _with The Markup, a nonprofit, investigative newsroom that challenges technology to serve the public good. Sign up for its newsletters_ _here__._

Less than 10 days after the Robb Elementary School shooting in Uvalde, Texas, in May 2022, Axon Enterprises CEO Rick Smith announced the company had formally started developing Taser-equipped drones. The technology, Smith argued, could potentially save lives during mass shootings by incapacitating active shooters within seconds.

For Axon, which changed its name from Taser in 2017, the concept seemed a sensible next step for stakeholders who share Axon’s public safety mission, Smith said on the company’s site.

“In brief,” he wrote, “non-lethal drones can be installed in schools and other venues and play the same role that sprinklers and other fire suppression tools do for firefighters: Preventing a catastrophic event, or at least mitigating its worst effects.”

Elsewhere, however, the announcement roused significant concern. Only a few weeks before Smith’s announcement, a majority of the members of Axon’s AI Ethics Board—which consisted of a dozen academics, attorneys, activists, and former law enforcement officials—recommended the company not move forward with a pilot study of Taser-armed drones, then called Project ION. The board had spent more than a year considering the Taser-equipped drone project but had never considered any use case in which it would be a solution to mass shootings.

Advisory board members told The Markup that Smith’s announcement was unexpected and made without consultation or input from the ethics body the company had worked relatively well with for the previous four years. In the past, the board’s work prompted Axon to ban facial recognition on its body cameras out of concern the technology could not be responsibly rolled out.

“I begged Rick not to go public with the weaponized drone plan without consulting with the board, as our operating principles required,” Barry Friedman, founder of NYU Law School’s Policing Project and former Axon ethics advisory board chair, told The Markup.

Within a week of the announcement, nine of Axon’s 12-member ethics board resigned, saying in a joint letter that they had “lost faith in Axon’s ability to be a responsible partner.”

“Although we all joined this Board understanding that we are advisory only—and have seen Axon reject our advice on some prior occasions—rushing ahead to embrace use of surveillance-enabled, Taser-equipped drones, especially when its Board was urging against unnecessarily precipitate action, is more than any of us can abide,” the exiting members wrote then.

An illustration of a Taser-equipped drone, taken from Axon’s patent application.Illustration: Axon

In the wake of the board’s dissolution, Axon halted its Taser-drone program temporarily. Former board members, meanwhile, continued to speak out against the company’s efforts. The group released a report in January 2023 criticizing company leaders for “trading on the tragic shootings which had just occurred in Uvalde and Buffalo.” The report included a number of recommendations about Taser-drone technology, including the need for accuracy and safety thresholds, as well as local lawmakers’ approval and internal department policies governing the drones’ use. Drones deploying force should never be autonomous, either, the ex-board members recommended—a human should make that decision.

The former members also noted that in addition to the physical risk of injury or death posed by an electroshock weapon, the proposed devices would rely on surveillance systems that would be triggered by the sound of gunshots, posing privacy and accuracy risks. Additional surveillance in schools might also lead to increased disciplinary action, even for minor offenses, they said. There’s potential for disparate, racist impact here, too, the former members pointed out, saying, “Black students are four times more likely to attend a school with a high level of surveillance.”

Weaponized drones are also vulnerable to misuse and might increase how frequently force is used, the experts said in their report. “A growing literature on military use of drones notes the unique characteristics of remote use of force—humans appear as figures on a computer screen, and decisions to use force often are made by teams rather than by a single individual,” the experts wrote. This “could lead to dehumanization of individuals targeted by the drone and could diminish operators’ sense of personal moral culpability for their decisions, leading to increased use of force.”

However, Axon, which did not reply to a request for comment by press time, still appears to be moving forward with its armed drone plan.

“On a longer time horizon, Axon sees opportunities to explore how robotics can expand to include less-lethal robotic payloads and operations,” company leadership said in a statement on its website from April of this year. “While this is still in early concept, we believe with ample research, ethical development and identifying the most amenable use cases, this capability can positively contribute to the future of public safety.”

Then, in July, the company acquired Sky-Hero, a company based in Belgium that manufactures drones and unmanned ground vehicles. Sky-Hero has already developed so-called “distraction” technology for some of its drones and rovers that produces the same sound pressure levels as a semiautomatic rifle, it said in the description of a YouTube video demonstrating the tech, acting as a “true non-lethal flashbang.”

In interviews with The Markup, former ethics advisory board members expressed concern about the company’s plans to continue developing weaponized drone technology.

“Vendors like Axon are selling products to public agencies for the benefit of the public, and I think they have a responsibility to consider the harms their products might cause and to try to mitigate those harms,” said Max Isaacs, a senior staff attorney at the Policing Project who worked with the board.

“Everyone deserves public safety,” Isaacs added. “Everyone wants more public safety. That’s not the question. The question is, when companies sell these products claiming all of these public safety benefits, have they proven those benefits? Has there been any independent testing? Do we know that these products are making us safer? Oftentimes the answer is no.”

Isaacs and others The Markup interviewed noted that ethics boards are an imperfect patch for the regulatory void around evolving technology like drones and artificial intelligence.

“The fact that we’re relying on companies to set up these advisory boards in order to address the harms of their products is itself a very problematic concept to me,” Isaacs said.

Though Axon leadership says it is committed to the “responsible” development of new technology, it’s not clear whether the company is still consulting with ethics experts on the plan.

The Markup found that mentions of the former AI ethics board appear to have been removed from the company’s website, including the board’s once-public recommendations and reports. The webpage axon.com/ethics, where the former board’s governing principles and work was hosted, now sends searchers to a letter from Smith announcing the company would pause its Taser-drone plan. In September 2022, the company unveiled its Ethics and Equity Advisory Council, a panel of academics and community leaders who advise Axon on “a limited number of products per year,” according to the company. Axon says the body is independent, but unlike the former ethics board, it is led by an Axon executive vice president.

The company declined to make members of this group available for interviews. Public reports from the body are not available on its website, nor is a copy of its operating principles.

Because ethics is both subjective and not legally binding, they can readily be trumped by capitalist imperatives, said Ryan Calo, a professor at the University of Washington School of Law and a former member of the ethics board at Axon.

“Ethics is important,” said Calo. “Ethics is good. But law needs to be a backstop. Law is the place where society goes to decide what’s forbidden and what’s required.”

The End of Killing?

The Arizona-based company has not shipped any Taser-equipped drones to any client, it said in a 2022–23 annual report, the most recent available. It also pledged to never build “lethal” drones.

Axon, which launched in 1993, is a premier player in the law enforcement and military technology space. Technology developed by the company, which makes Taser devices and body-worn cameras, is used by more than 95 percent of state and local law enforcement agencies in the United States, it claims in investor reports. Axon also owns the evidence.com platform, a cloud-based evidence management system for police officers and what the company calls “the world’s leading repository of law enforcement data.” Regarding Taser drones, Axon maintains they should be developed. “We believe there is no organization in the world better suited to develop it the right way,” its most recent investor report reads.

Company leadership expects to generate more than $1.5 billion in revenue in 2023, it said in an August investor statement. And by 2025, Rick Smith has set a target of reaching $2 billion. According to its own reports, Axon, which went public in 2001, has generated “over $15 billion in wealth” for its shareholders.

Still, the plan to arm drones with Tasers was not universally well-received by Axon’s shareholders, some of whom criticized the company for Smith’s announcement about the weaponized drones. A shareholder proposal submitted by the Jubitz Family Foundation, a Portland, Oregon–based foundation that promotes nonviolent alternatives to conflict, encouraged shareholders to vote to discontinue developing these drones.

“Axon proposed using AI surveillance, algorithmic predictors, and virtual reality simulations to stop mass shootings,” the proposal, which was included as part of the company’s 2022–23 annual report, reads. “Axon did not seek meaningful input from its in-house Community Advisory Coalition, AI Ethics Board, or Vice President of Community Impact prior to the announcement.”

After the ethics board’s resignations last year, “Axon has now replaced both the Community Advisory Coalition and the AI Ethics Board with a new advisory council, which Smith still does not commit to heeding,” the foundation added in its proposal.

“The rollout of this proposal demonstrates a tremendous failure of management’s self-governance procedures,” the Foundation wrote, and risked not only harming children psychologically and physically, but possible litigation and reputational damage.

The Jubitz Family Foundation did not respond to a request for comment on the proposal from The Markup.

In a lengthy response to the Jubitz proposal, Axon said robotic security could save lives, slashing gun-related deaths by presenting police with longer-range, remotely operable weapons.

“Axon is working to reduce violence and displace lethal uses of force with less-lethal alternatives that can save—rather than take—lives,” the company said.

“Based on our analysis of _The Washington Post_’s data set of fatal officer-involved shootings, we estimate that a more effective, longer-range handheld Taser device has the potential to reduce fatal officer-involved shootings by around 40 percent,” the company said. “When we run this same analysis looking at instances where police could have utilized a less-lethal capable drone, we estimate that a drone could likely have been used instead of lethal force in 57 percent of these fatal shootings. When we combine an advanced handheld Taser device together with remotely operated drone and robotic capabilities, we estimate that up to 72 percent of fatal shootings might be averted.” (The company did not share information about its analysis in response to questions about it.)

While Axon technology is used by major police departments and federal agencies, including the New York Police Department, the Los Angeles Police Department, the US Department of Homeland Security, and the Departments of Defense and Justice, according to the company, there isn’t proof that the products are solving the problem of police violence. According to the _Washington Post_ database of fatal police shootings, the number of such shootings was higher in 2022 than it was in any of the previous seven years tracked. And recently, some police unions have argued they should be paid more just to use body cameras, a barrier to critical transparency even where these tools are available.

Axon did not respond to questions in time for publication about when taser-equipped drones might arrive on the market, whether its new Equity and Ethics Advisory Council is advising it on the weaponized drone program, or whether the company has received any demand from school districts for these products. However, some insight into CEO Smith’s vision for this kind of drone technology can be gleaned from his 2019 book _The End of Killing_.

In a chapter of the book on school safety, Smith presents readers with a fictitious scenario about a day care center shooting, a tragedy averted because of a Taser-equipped drone installed in the room and activated by a “AI algorithm … designed to constantly monitor for potential firearm discharge sounds, not all that different from the iPhones of millions of people around the world that are awakened by the ‘Hey Siri’ sound pattern,” Smith writes.

An algorithm, Smith writes, would subsequently calculate the direction of the sound and, combined with a panic-alert signal system, trigger the drop of a small drone within a second. A “computer vision algorithm” on the drone would detect muzzle flashes. Sensing a probable weapon, he added, police could remotely deploy the Taser-armed drone, shooting electrical impulses “designed to paralyze the human nervous system” toward the shooter. All of this could take place within two seconds of gunfire.

Axon's Ethics Board Resigned Over Taser-Armed Drones. Then the Company Bought a Military Drone Maker

Authorities have sanctioned 11 alleged members of the cybercriminal groups, while the US Justice Department unsealed three federal indictments against nine people accused of being members.

The United States Department of Treasury and United Kingdom Foreign Office announced today that they have sanctioned 11 people for their alleged involvement in the Trickbot cybercriminal gang. The US Department of Justice also unsealed indictments against nine people whom it says are connected to Trickbot and its sibling organization Conti. Seven of those nine also appear on today's sanctions list.

US and UK law enforcement working with officials around the world have made a concerted effort in recent years to deter cybercrime—particularly ransomware attacks and those launched by Russia-based actors. And Trickbot, a notorious and prolific gang, has repeatedly been a specific target of these actions. In February, the US and UK announced sanctions against seven alleged Trickbot actors and an indictment against them.

The new round of censures includes alleged Trickbot members who are accused of acting as coders and administrators for the group, as well as senior staff, the developer team lead, and a human resources and finance manager. The sanctions also name Trickbot's alleged head of testing for the gang's malware and technical infrastructure. This individual, Maksim Galochkin, goes by the handle Bentley, among others. WIRED identified Galochkin last week as part of an extensive investigation into Trickbot and its operations. 

The Department of Justice announced three indictments today that include Galochkin. One in the Northern District of Ohio, filed on June 15, charges him and 10 other alleged Trickbot members with “conspiring to use the Trickbot malware to steal money and personal and confidential information from unsuspecting victims, including businesses and financial institutions located in the United States and around the world, beginning in November 2015.” This timeline means that the charges essentially relate to all Trickbot activity going back to the group's inception. 

An indictment from the Middle District of Tennessee, filed on June 12, charges Galochkin and three others with use of the Conti ransomware in attacks targeting “businesses, nonprofits, and governments in the United States” between 2020 and June 2022. And an indictment in the Southern District of California, filed on June 14, charges Galochkin in connection with the May 1, 2021, Conti ransomware attack on Scripps Health.

“Today’s announcement shows our ongoing commitment to bringing the most heinous cyber criminals to justice—those who have devoted themselves to inflicting harm on the American public, our hospitals, schools, and businesses,” FBI director Christopher Wray said in a statement on Thursday. “Cyber criminals know that we will use every lawful tool at our disposal to identify them, tirelessly pursue them, and disrupt their criminal activity. We, alongside our federal and international partners, will continue to impose costs through joint operations no matter where these criminals may attempt to hide.”

It has been difficult for global law enforcement to make progress on deterring cybercrminal activity, especially when actors are based in countries like Russia that allow them to operate with impunity. But independent researchers say that imposing public accountability does have impacts on the individuals as well as the broader criminal landscape.

Cybercriminals “often think they can conduct cyberattacks against corporations and individuals under anonymity,” says Landon Winkelvoss, vice president of research for the digital intelligence firm Nisos, which conducted a detailed investigation of Bentley's real-world identity at WIRED's request. But “they all make mistakes and the very nature of their crimes requires that their digital footprint is in the wild."

Winkelvoss notes that while cybercriminals have systematized strategies for maintaining their operational security and staying out of the limelight, their efforts to remain invisible are far from foolproof.

“Reusing command and control infrastructure servers and selectors like emails addresses and phone numbers is often the quickest return on their investment,” Winkelvoss says. "Unfortunately for them, this makes their unmasking relatively straightforward, especially when law enforcement and private industry \[have\] more publicly available data than they do.”

US and UK Mount Aggressive Crackdown on Trickbot and Conti Ransomware Gangs

And the first case on the docket may well be Russia’s cyberattacks against civilian critical infrastructure in Ukraine.

For years, some cybersecurity defenders and advocates have called for a kind of Geneva Convention for cyberwar, new international laws that would create clear consequences for anyone hacking civilian critical infrastructure, like power grids, banks, and hospitals. Now the lead prosecutor of the International Criminal Court at the Hague has made it clear that he intends to enforce those consequences—no new Geneva Convention required. Instead, he has explicitly stated for the first time that the Hague will investigate and prosecute any hacking crimes that violate existing international law, just as it does for war crimes committed in the physical world.

In a little-noticed article released last month in the quarterly publication _Foreign Policy Analytics_, the International Criminal Court’s lead prosecutor, Karim Khan, spelled out that new commitment: His office will investigate cybercrimes that potentially violate the Rome Statute, the treaty that defines the court’s authority to prosecute illegal acts, including war crimes, crimes against humanity, and genocide.

“Cyber warfare does not play out in the abstract. Rather, it can have a profound impact on people’s lives,” Khan writes. “Attempts to impact critical infrastructure such as medical facilities or control systems for power generation may result in immediate consequences for many, particularly the most vulnerable. Consequently, as part of its investigations, my Office will collect and review evidence of such conduct.”

When WIRED reached out to the International Criminal Court, a spokesperson for the office of the prosecutor confirmed that this is now the office’s official stance. “The Office considers that, in appropriate circumstances, conduct in cyberspace may potentially amount to war crimes, crimes against humanity, genocide, and/or the crime of aggression,” the spokesperson writes, “and that such conduct may potentially be prosecuted before the Court where the case is sufficiently grave.”

Neither Khan’s article nor his office’s statement to WIRED mention Russia or Ukraine. But the new statement of the ICC prosecutor’s intent to investigate and prosecute hacking crimes comes in the midst of growing international focus on Russia’s cyberattacks targeting Ukraine both before and after its full-blown invasion of its neighbor in early 2022. In March of last year, the Human Rights Center at UC Berkeley’s School of Law sent a formal request to the ICC prosecutor’s office urging it to consider war crime prosecutions of Russian hackers for their cyberattacks in Ukraine—even as the prosecutors continued to gather evidence of more traditional, physical war crimes that Russia has carried out in its invasion.

In the Berkeley Human Rights Center’s request, formally known as an Article 15 document, the Human Rights Center focused on cyberattacks carried out by a Russian group known as Sandworm, a unit within Russia’s GRU military intelligence agency. Since 2014, the GRU and Sandworm, in particular, have carried out a series of cyberwar attacks against civilian critical infrastructure in Ukraine beyond anything seen in the history of the internet. Their brazen hacking has ranged from targeting Ukrainian electric utilities and triggering the only two blackouts ever caused by cyberattacks to the release of the data-destroying NotPetya malware that spread from Ukraine to the rest of the world and inflicted more than $10 billion in damage, including to hospital networks in both Ukraine and the United States.

Though the Berkeley group’s submission initially focused on Sandworm’s 2015 and 2016 attacks on Ukraine’s power grid as the clearest example of cyberattacks with physical effects comparable to those of traditional warfare, it later expanded its argument to include Sandworm’s NotPetya cyberattack, as well as a third attempt by the hackers to sabotage Ukraine’s power grid and another cyberattack on the Viasat satellite modem network used by Ukraine’s military, which caused outages of the satellite modems across Europe.

The fact that Khan doesn’t explicitly mention Russia in his article doesn’t actually mean he’s shying away from investigating war crimes carried out by Sandworm or other Russians involved in the Ukraine attacks, says Lindsay Freeman, the Human Rights Center’s director of technology, law, and policy. Instead, she sees the article as a broader statement that hacking activities that violate international law will be considered as part of any investigation the prosecutors carry out. “The fact that he’s not just saying that he’s going to do this in Ukraine, that he’s going to do this in _all_ investigations is really important,” says Freeman. “Seeing that this is the reality of cyberwar now and this is something they, as an office, have to investigate in every single case—that goes beyond what we were pushing for, and it’s a really important and powerful move.”

Ukraine’s government, meanwhile, has already opened its own investigation into Russian war crimes carried out via cyberattacks. Aside from charging any Russian hackers or their superiors in its own court system, evidence from that investigation could now also be provided to the ICC’s prosecutors to aid any case that the Hague prosecutors bring against Russia.

Six of Sandworm’s hackers are already facing an indictment in the United States for hacking crimes related to their cyberattacks targeting Ukraine, as well as the network of the 2018 Winter Olympics in Pyeongchang, Korea. But Freeman points out that charges against Russian hackers at the Hague would have a broader effect: 123 countries are parties to the Rome Statute, and so have agreed to help detain and extradite convicted war criminals. That includes some countries that don’t have extradition treaties with the United States, such as Switzerland and Ecuador. The Hague’s remit also extends not only to the hands-on-keyboards hackers themselves, Freeman notes, but to the command structure above those hackers, opening the possibility of new charges against higher-level officers within Russia’s military or even Russian president Vladimir Putin himself.

In terms of legal precedents, Khan’s statement that the Hague prosecutor’s office will now consider hacking a potential violation of international law is “novel but not surprising,” says Bobby Chesney, director of the Strauss Center for International Security and Law at the University of Texas Law School. “I don’t think anyone who’s serious about international law would dispute that there are at least some circumstances in which intentional harm to civilians can be carried out through cyber means in a way that qualifies as an attack and therefore a violation” of the Rome Statute’s principle that combatants distinguish between civilian and military targets, Chesney says.

Chesney says he was more intrigued to see other parts of Khan’s article that mention disinformation as a separate area of concern and “gray zone” tactics that “operate in the area between war and peace.” As a precedent for sources of disinformation being charged under international law, Chesney points to rare incidents, like radio journalists who were convicted in an international criminal tribunal for helping incite genocide in Rwanda in 1994. Investigating or charging hacking that occurs outside the context of war as a violation of international law, as Khan’s article seems to suggest, would be newer territory.

But the Human Rights Center’s Freeman argues that given the ICC prosecutor’s limited resources and discretion to choose which cases to prosecute, any commitment to examine and potentially charge cyberwar crimes is a historic moment. “Just looking at how cyber is being used in war, and that \[Khan\] sees it in his remit and something worth investigating as a priority within his power of discretion, I think is incredibly important,” she says.

Khan clearly sees the stakes as equally high: He ends his article by quoting (or perhaps misquoting) Albert Einstein’s stated fear that “technology would exceed our humanity.”

“Undoubtedly, we shall be tested,” Khan writes. “But through our common efforts—and above all the belief that we can mobilize the law on these new front lines to deliver justice—we may collectively ensure that a more humane world is forged. The ICC will play its part, now and in years to come.”

The International Criminal Court Will Now Prosecute Cyberwar Crimes

Here's how to request that your personal information not be used to train Meta's AI model. "Request" is the operative word here.

As Meta, the company behind Facebook, continues to develop its generative artificial intelligence tools, you can now request the removal of some of the personal data the company uses to train its AI model. There are a ton of caveats, though.

Earlier this year, CEO Mark Zuckerberg announced plans to build a range of AI features into Meta’s platforms like Facebook, Instagram, and WhatsApp. Despite the popularity of generative AI in Silicon Valley, murky legal questions remain for the technology, and many people are anxious about its rapid advancement.

Want to stop Meta from using all of your info to improve its AI? The company added a new form to one of its help centers titled “Generative AI Data Subject Rights at Facebook.” With this form, you can request that Meta give you access to the third-party data it uses for AI development and that the personal information is deleted. The operative word here is “request.” There’s no guarantee from the company that it’ll delete it, or that it’ll provide you with the information you’re asking for, even if it’s yours.

It’s important to point out that this form does not pertain to the gobs of personal information Meta has already collected from you on its platforms; it only applies to outside data the company may bring in to beef up its generative AI. This outside data could include stuff that’s elsewhere on the internet as well as data purchased from third-party data brokers.

You don’t need to be signed in to a Facebook account to submit the opt-out request. All that Facebook requires is your country of residence, full name, and email address. In my experience, the website was quite glitchy on mobile, and it was easier to fill out the form on my desktop. It took the company over 24 hours to send a basic confirmation email stating it was “reviewing \[my\] request.”

For those in the US, it’s unclear if anything happens at all when you fill out this form. Data privacy laws protecting UK residents make this form more worthwhile for people who live in the UK. “Depending on where people live, they may be able to exercise their data subject rights and object to certain data being used to train our AI models,” said Thomas Richards, a Meta spokesperson, to Gizmodo. Meta did not respond to multiple requests from WIRED to comment on this story.

A company blog post about Meta’s use of personal info to train its AI goes just a little further in detail about its different approaches to personal data collection, but specifics are scant: “In the European region and United Kingdom, we rely on the basis of legitimate interests to collect and process any personal information included in these publicly available and licensed sources to train our generative AI models. For other jurisdictions where applicable, we rely on an adequate legal basis to collect and process this data.”

This opt-out form comes just a few months after European regulators struck Meta with a $1.3 billion fine for misusing data that originated in the UK.

Facebook Trains Its AI on Your Data. Opting Out May Be Futile

After leaving many questions unanswered, a new post mortem from Microsoft explains the series of slip-ups that allowed attackers to steal and abuse a valuable cryptographic key.

Microsoft said in June that a China-backed hacking group had stolen a cryptographic key from the company's systems. This key allowed the attackers to access cloud-based Outlook email systems for 25 organizations, including multiple United States government agencies. At the time of the disclosure, however, Microsoft did not explain how the hackers were able to compromise such a sensitive and highly guarded key, or how they were able to use the key to move between consumer- and enterprise-tier systems. But a new post mortem published by the company on Wednesday explains a chain of slip-ups and oversights that allowed the improbable attack.

Such cryptographic keys are significant in cloud infrastructure because they are used to generate authentication “tokens” that prove a user’s identity for accessing data and services. Microsoft says it stores these sensitive keys in an isolated and strictly access-controlled “production environment.” But during a particular system crash in April 2021, the key in question was an incidental stowaway in a cache of data that crossed out of the protected zone.

“All the best hacks are deaths by 1,000 paper cuts, not something where you exploit a single vulnerability and then get all the goods,” says Jake Williams, a former US National Security Agency hacker who is now on the faculty of the Institute for Applied Network Security.

After the fateful crash of a consumer signing system, the cryptographic key ended up in an automatically generated “crash dump” of data about what had happened. Microsoft's systems are meant to be designed so signing keys and other sensitive data doesn't end up in crash dumps, but this key slipped through because of a bug. Worse still, the systems built to detect errant data in crash dumps failed to flag the cryptographic key.

With the crash dump seemingly vetted and cleared, it was moved from the production environment to a Microsoft “debugging environment,” a sort of triage and review area connected to the company's regular corporate network. Once again though, a scan designed to spot the accidental inclusion of credentials failed to detect the key's presence in the data.

Sometime after all of this occurred in April 2021, the Chinese espionage group, which Microsoft calls Storm-0558, compromised the corporate account of a Microsoft engineer. With this account, the attackers could access the debugging environment where the ill-fated crash dump and key were stored. Microsoft says it no longer has logs from this era that directly show the compromised account exfiltrating the crash dump, “but this was the most probable mechanism by which the actor acquired the key.” Armed with this crucial discovery, the attackers were able to start generating legitimate Microsoft account access tokens.

Another unanswered question about the incident had been how the attackers used a cryptographic key from the crash log of a consumer signing system to infiltrate the enterprise email accounts of organizations like government agencies. Microsoft said on Wednesday that this was possible because of a flaw related to an application programming interface that the company had provided to help customer systems cryptographically validate signatures. The API had not been fully updated with libraries that would validate whether a system should accept tokens signed with consumer keys or enterprise keys, and as a result, many systems could be tricked into accepting either.

The company says it has fixed all of the bugs and lapses that cumulatively exposed the key in the debugging environment and allowed it to sign tokens that would be accepted by enterprise systems. But the recap does not describe how attackers compromised the engineer's corporate account, and Microsoft did not immediately respond to WIRED's request for comment about how the account breach occurred. This is important information for fully understanding how the attack played out, independent security researcher Adrian Sanabria says. He adds, too, that the fact Microsoft kept limited logs during this time period is significant. As part of its response to the Storm-0558 hacking spree overall, the company said in July that it would expand the cloud logging capabilities that it offers for free.

“It's particularly notable, because one of the complaints about Microsoft is that they don't set up their own customers for security success,” Sanabria says. “Logs disabled by default, security features are an add-on requiring additional spending, or more premium licenses. It appears they themselves got bit by this practice.”

As Williams from The Institute for Applied Network Security points out, organizations like Microsoft must face highly motivated and well-resourced attackers who are unusually capable of capitalizing on the most esoteric or improbable mistakes. He says that from reading Microsoft's latest updates on the situation, he is more sympathetic to why the situation played out the way it did.

“You'll only hear about highly complex hacks like this in an environment like Microsoft's,” he says. “In any other organization the security is relatively so weak that a hack doesn't need to be complex. And even when environments are pretty secure, they often lack the telemetry—along with the retention—needed to investigate something like this. Microsoft is a rare organization that has both. Most organizations wouldn't even store logs like this for a few months so I'm impressed that they had as much telemetry as they did."

Source

Wired