The number of alleged hacks targeting the customers of cloud storage firm Snowflake appears to be snowballing into one of the biggest data breaches of all time.

A hack against customers of the cloud storage company Snowflake looks like it may turn into one of the biggest-ever data breaches. Last week, Snowflake, which allows companies to store huge datasets on its servers, revealed that criminal hackers had been attempting to access its customers’ accounts using stolen login details. Data breaches targeting Ticketmaster and Santander have been linked to the attacks.

In the days since Snowflake first said a “limited number” of customer accounts had been accessed, however, cybercriminals have publicly claimed to be selling stolen data from two other major firms and alleged the information was taken from Snowflake accounts. At the same time, TechCrunch has reported that hundreds of Snowflake customer passwords have been found online and are accessible to cybercriminals.

Amid the claims, there remains uncertainty about the scope and scale of the attempted attack against Snowflake customers, who the attackers may be, and how an attack tool callously named “rapeflake” operates. It also highlights the growth in the use of infostealer malware in recent years and underscores the need for third-party software providers and companies to turn on multifactor authentication to reduce the chances of accounts being compromised.

**Snowballing**

A large proportion of the Snowflake drama has, so far, played out on the notorious cybercrime marketplace BreachForums. The FBI seized the forum in mid May, but another version quickly appeared, and its owners, hacker group ShinyHunters, claimed to be selling 560 million records from Ticketmaster and 30 million from Santander. Both companies have said they have suffered from data breaches, with Ticketmaster directly linking the incident to Snowflake while Santander said it had seen unauthorized access to one of its databases “hosted by a third-party provider.” Neither company has confirmed the size of the breaches.

In recent days, a BreachForums account going by the handle Sp1d3r has posted two more companies whose data it claims is related to the Snowflake incident: automotive giant Advance Auto Parts, which Sp1d3r claims to have 380 million customer details from, and financial services company LendingTree and subsidiary QuoteWizard, from which it claims to have data linked to 190 million people.

Some Advance Auto Parts staff and customer email addresses listed in sample data by the hacker appear to be legitimate accounts; emails WIRED sent to those addresses did not bounce nor were they otherwise rejected. BleepingComputer reports it has verified customer data from Advance Auto Parts.

“We are aware of reports that Advance may be involved in a security incident related to Snowflake,” Darryl Carr, a spokesperson from the company, tells WIRED. “We are investigating the matter and do not have further information to share at this time. We have not experienced any impact to our operations or systems.”

LendingTree has not responded to multiple requests from WIRED about the alleged breaches sent in the past few days. Neither LendingTree nor Advance Auto Parts has filed breach notifications with the Securities and Exchange Commission at the time of writing. Both companies have been listed by Snowflake as customers previously.

Since Snowflake acknowledged that accounts had been targeted, it has provided some more information about the incident. Brad Jones, Snowflake’s chief information security officer, said in a blog post that threat actors used login details to accounts that had been “purchased or obtained through infostealing malware,” which is designed to pull usernames and passwords from devices that have been compromised. The incident appears to be a “targeted campaign directed at users with single-factor authentication,” Jones added.

Jones’ post said Snowflake, alongside cybersecurity companies CrowdStrike and Mandiant, which it employed to investigate the incident, did not find evidence showing the attack was “caused by compromised credentials of current or former Snowflake personnel.” However, it has found one former employee's demo accounts were accessed, claiming they did not contain sensitive data.

When asked about potential breaches of specific companies’ data, a Snowflake person pointed to Jones’ statement: “We have not identified evidence suggesting this activity was caused by a vulnerability, misconfiguration, or breach of Snowflake’s platform.” The company did not provide an on-record comment clarifying what was meant by a “breach.” (Security company Hudson Rock said it removed a research post including various unverified claims about the Snowflake incident after receiving a legal letter from Snowflake).

The US Cybersecurity and Infrastructure Security Agency has issued an alert about the Snowflake incident, while Australia's Cyber Security Center said it is “aware of successful compromises of several companies utilizing Snowflake environments.”

**Unclear Origins**

Little is known about the Sp1d3r account advertising data on BreachForums, and it is not clear whether ShinyHunters obtained the data it was selling from another source or directly from victims’ Snowflake accounts—information about a Ticketmaster and Santander breach was originally posted on another cybercrime forum by a new user called SpidermanData.

The Sp1d3r account posted on BreachForums that the 2 terabytes of alleged LendingTree and QuoteWizard data was for sale for $2 million; while 3 TB of data allegedly from Advance Auto Parts would cost someone $1.5 million. “The price set by the threat actor appears extremely high for a typical listing posted to BreachForums,” says Chris Morgan, a senior cyber-threat intelligence analyst at security firm ReliaQuest.

Morgan says the legitimacy of Sp1d3r is not clear; however, he points out there is a nod to teenage hacking group Scattered Spider. “Interestingly, the threat actor's profile picture is taken from an article referencing the threat group Scattered Spider, although it is unclear whether this is to make an intentional association with the threat group.”

While the exact source of the alleged data breaches is unclear, the incident highlights how interconnected companies can be when relying on products and services from third-party providers. “I think a lot of this is just a recognition of how interdependent these services now are and how hard it is to control the security posture of third parties,” security researcher Tory Hunt told WIRED when the incidents first emerged.

As part of its response to the attacks, Snowflake has told all customers to make sure they enforce multifactor authentication on all accounts and allow traffic only from authorized users or locations. Companies that have been impacted should also reset their Snowflake login credentials. Enabling multifactor authentication vastly reduces the chances that online accounts will be compromised. As mentioned, TechCrunch reported this week that it has seen “hundreds of alleged Snowflake customer credentials'' taken by infostealing malware from computers of people who have accessed Snowflake accounts.

In recent years, coinciding with more people working from home since the Covid-19 pandemic, there has been a rise in the use of infostealer malware. “Infostealers have become more popular because they’re in high demand and pretty easy to create,” says Ian Gray, the vice president of intelligence at security company Flashpoint. Hackers have been seen to be copying or modifying existing infostealers and selling them on for as little as $10 for all the login details, cookies, files, and more from one infected device.

“This malware can be delivered in different ways and targets sensitive info like browser data (cookies and credentials), credit cards, and crypto wallets,” Gray says. “Hackers might comb through the logs for enterprise credentials to break into accounts without permission.”

The Snowflake Attack May Be Turning Into One of the Largest Data Breaches Ever

ZeroMark wants to build a system that will let soldiers easily shoot a drone out of the sky with the weapons they’re already carrying—and venture capital firm a16z is betting the startup can pull it off.

Drones have changed war. Small, cheap, and deadly robots buzz in the skies high above the world’s battlefields, taking pictures and dropping explosives. They’re hard to counter. ZeroMark, a defense startup based in the United States, thinks it has a solution. It wants to turn the rifles of frontline soldiers into “handheld Iron Domes.”

The idea is simple: Make it easier to shoot a drone out of the sky with a bullet. The problem is that drones are fast and maneuverable, making them hard for even a skilled marksman to hit. ZeroMark’s system would add aim assistance to existing rifles, ostensibly helping soldiers put a bullet in just the right place.

“We’re mostly a software company,” ZeroMark CEO Joel Anderson tells WIRED. He says that the way it works is by placing a sensor on the rail mount at the front of a rifle, the same place you might put a scope. The sensor interacts with an actuator either in the stock or the foregrip of the rifle that makes adjustments to the soldier’s aim while they’re pointing the rifle at a target.

A soldier beset by a drone would point their rifle at the target, turn on the system, and let the actuators solidify their aim before pulling the trigger. “So there’s a machine perception, computer vision component. We use lidar and electro-optical sensors to detect drones, classify them, and determine what they’re doing,” Anderson says. “The part that is ballistics is actually quite trivial … It’s numerical regression, it’s ballistic physics.”

According to Anderson, ZeroMarks’ system is able to do things a human can’t. “For them to be able to calculate things like the bullet drop and trajectory and windage … It’s a very difficult thing to do for a person, but for a computer, it’s pretty easy,” he says. “And so we predetermined where the shot needs to land so that when they pull the trigger, it’s going to have a high likelihood of intersecting the path of the drone.”

ZeroMark makes a tantalizing pitch—one so attractive that venture capital firm Andreesen Horowitz invested $7 million in the project. The reasons why are obvious for anyone paying attention to modern war. Cheap and deadly flying robots define the conflict between Russia and Ukraine. Every month, both sides send thousands of small drones to drop explosives, take pictures, and generate propaganda.

With the world’s militaries looking for a way to fight back, counter-drone systems are a growth industry. There are hundreds of solutions, many of them not worth the PowerPoint slide they’re pitched from.

Can a machine-learning aim-assist system like what ZeroMark is pitching work? It remains to be seen. According to Anderson, ZeroMark isn’t on the battlefield anywhere, but the company has “partners in Ukraine that are doing evaluations. We’re hoping to change that by the end of the summer.”

There’s good reason to be skeptical. “I’d love a demonstration. If it works, show us. Till that happens, there are a lot of question marks around a technology like this,” Arthur Holland Michel, a counter-drone expert and senior fellow at the Carnegie Council for Ethics in International Affairs, tells WIRED. “There’s the question of the inherent unpredictability and brittleness of machine-learning-based systems that are trained on data that is, at best, only a small slice of what the system is likely to encounter in the field.”

Anderson says that ZeroMark’s training data is built from “a variety of videos and drone behaviors that have been synthesized into different kinds of data sets and structures. But it’s mostly empirical information that’s coming out of places like Ukraine.”

Michel also contends that the physics, which Anderson says are simple, are actually quite hard. ZeroMark’s pitch is that it will help soldiers knock a fast-moving object out of the sky with a bullet. “And that is very difficult,” Michel says. “It’s a very difficult equation. People have been trying to shoot drones out of the sky \[for\] as long as there have been drones in the sky. And it’s difficult, even when you have a drone that is not trying to avoid small arms fire.”

That doesn’t mean ZeroMark doesn’t work—just that it’s good to remain skeptical in the face of bold claims from a new company promising to save lives. “The only truly trustworthy metric of whether a counter-drone system works is if it gets used widely in the field—if militaries don’t just buy three of them, they buy thousands of them,” Michel says. “Until the Pentagon buys 10,000, or 5,000, or even 1,000, it’s hard to say, and a little skepticism is very much merited.”

The Lords of Silicon Valley Are Thrilled to Present a ‘Handheld Iron Dome’

Eliot Higgins and his 28,000 forensic foot soldiers at Bellingcat have kept a miraculous nose for truth—and a sharp sense of its limits—in Gaza, Ukraine, and everywhere else atrocities hide online.

How to Lead an Army of Digital Sleuths in the Age of AI

A WIRED investigation, based on more than 22 million flight coordinates, reveals the complicated truth about the first full-blown police drone program in the US—and why your city could be next.

The Age of the Drone Police Is Here

Wyoming’s secretary of state has proposed ways of “preventing fraud and abuse of corporate filings by commercial registered agents” in the aftermath of the scheme’s exposure.

For years, the Federal Bureau of Investigation has been unraveling what it asserts is a scam perpetrated by agents of North Korea, which used fake companies employing real IT workers to funnel money back to the regime’s military.

An American company played a key role in creating shell companies used as part of the scheme, a WIRED review of public records shows. Elected officials are now contemplating addressing loopholes in business-registration law that the scheme exposed.

In May, Wyoming secretary of state Chuck Gray revoked the business licenses of three companies linked to the North Korean scam: Culture Box LLC, Next Nets LLC, and Blackish Tech LLC. Gray said his office made the decision after receiving information from the FBI and conducting an investigation.

“The communist, authoritarian Kim Jong Un regime has no place in Wyoming,” Gray said in a May press release.

The companies posed as legitimate operations where businesses could hire contract workers to perform IT solutions, complete with fake websites featuring smiling photos of apparent employees. The companies all had one thing in common: Their incorporation documents were filed by a company called Registered Agents Inc., which says its global headquarters is in Sheridan, Wyoming.

Registered Agents, which provides incorporation services in every US state, takes the practice of business privacy to the extreme, and regularly uses fake personae to file formation documents with state agencies, a WIRED investigation previously found.

Culture Box LLC, one of the companies that Gray and the FBI linked to North Korea, listed “Riley Park” as the name of a Registered Agents employee on documents submitted to the Wyoming secretary of state. Park, according to several former employees of Registered Agents, is a fake persona that the company regularly used to file incorporation documents.

In a statement provided to WIRED, Registered Agents wrote, “The Wyoming Secretary of State dissolved the entities and we initiated the 30-day process to resign as their agent in mid-May. Ours and Wyoming's processes to identify bad actors works. It strikes the best balance of individual privacy and business transparency supported by an entire ecosystem that cares about supporting entrepreneurs while rooting out the small percent of scammers.” The FBI’s St. Louis office, which led the investigation, did not respond to a request for comment.

The North Korean operation worked like this: Agents of the regime created fake companies purporting to be legitimate firms offering freelance IT services. Workers hired by North Koreans, or North Koreans themselves, would then perform legitimate contractor work, often using assumed identities.

In some instances, Americans would set up low-cost laptops with remote-access software, allowing North Korean workers to perform freelance IT work while appearing to use American IP addresses. The FBI referred to these Americans as “virtual assistants.”

The payments for the IT work were eventually funneled back to North Korea—where, the Department of Justice asserts, it was directed to the country’s Ministry of Defense and other agencies involved in WMD work. The scheme was so expansive that any company that hired freelance IT workers “more than likely” hired someone involved in the operation, according to FBI agent Jay Greenberg.

The shell companies created in Wyoming were used to hire virtual assistants and receive payments.

“I discovered that North Korean IT workers create and use domain names and limited liability companies (LLCs) in furtherance of their fraudulent activity and to mask their true identities as North Koreans. The LLCs are used to recruit ‘Virtual Assistants’ who can receive and ship devices needed for the North Korean IT workers as well as recruit and employ software developers from countries such as Pakistan, India, and China,” an FBI agent wrote in a May affidavit. “These LLCs are often registered in the United States through business registry services and sometimes use the identities of individuals who had a previous relationship with North Korean IT workers.”

The affidavit alleges that money from North Korean workers was used to purchase domain names for the IT front companies, in violation of sanctions laws. The domains were purchased using “payment service providers” with accounts belonging to the Wyoming companies.

In response to a request for comment from WIRED, the Wyoming secretary of state’s office said that it has “increased the number of complete, in-person audits of commercial registered agents, resulting in several ongoing investigations, as well as the issuance of findings and orders.”

The secretary of state has offered proposals to the Wyoming state legislature “aimed at preventing fraud and abuse of corporate filings by commercial registered agents, as ways to strengthen the Wyoming secretary of state's administrative authority to dissolve business entities controlled by foreign adversaries,” said Joe Rubino, the chief policy officer and general counsel at the Wyoming Secretary of State's Office.

An American Company Enabled a North Korean Scam That Raised Money for WMDs

TikTok has confirmed a “potential exploit” that is being used to go after accounts belonging to media organizations and celebrities, including CNN and Paris Hilton, through direct messages.

TikTok says it's currently taking steps to mitigate a cyberattack that's targeting a number of high-profile users through direct messages, in an attempt to hijack their accounts.

“We have taken measures to stop this attack and prevent it from happening in the future. We're working directly with affected account owners to restore access, if needed,” says Jason Grosse, a spokesperson for TikTok’s privacy and security team.

Grosse says TikTok is still investigating the attack and could not comment at this time about its scale or sophistication, describing the threat as merely a “potential exploit.”

TikTok’s acknowledgment followed a report on Tuesday claiming CNN's account had been temporarily breached last week. Citing an anonymous source at the news organization, Semafor reports that the breach did "not appear to be the result of someone gaining access from CNN's end." CNN did not immediately respond to WIRED’s request to comment.

Concerns over hacking attempts targeting news organizations in the US are particularly high given the impending presidential election this fall.

Forbes reported earlier in the day that the account of hotel heiress Paris Hilton was similarly affected, citing sources within the company. A source at TikTok tells WIRED that Hilton’s account was targeted but had not been compromised.

Security and privacy concerns around TikTok expand beyond cyberattacks by malicious actors. The company itself is fighting to remain available in the United States after US president Joe Biden signed a law in April that forces its parent company, China-based ByteDance, to sell TikTok or face a ban. TikTok and several users have sued the US government, claiming the law is unconstitutional on First Amendment grounds.

_Update 3:30 pm ET, June 4, 2024: A TikTok spokesperson tells_ WIRED _that the company has detected only two accounts compromised by the attack, including CNN’s._

TikTok Hack Targets ‘High-Profile’ Users via DMs

Windows Recall takes a screenshot every five seconds. Cybersecurity researchers say the system is simple to abuse—and one ethical hacker has already built a tool to show how easy it really is.

When Microsoft CEO Satya Nadella revealed the new Windows AI tool that can answer questions about your web browsing and laptop use, he said one of the “magical” things about it was that the data doesn’t leave your laptop; the Windows Recall system takes screenshots of your activity every five seconds and saves them on the device. But security experts say that data may not stay there for long.

Two weeks ahead of Recall’s launch on new Copilot+ PCs on June 18, security researchers have demonstrated how preview versions of the tool store the screenshots in an unencrypted database. The researchers say the data could easily be hoovered up by an attacker. And now, in a warning about how Recall could be abused by criminal hackers, Alex Hagenah, a cybersecurity strategist and ethical hacker, has released a demo tool that can automatically extract and display everything Recall records on a laptop.

Dubbed TotalRecall—yes, after the 1990 sci-fi film—the tool can pull all the information that Recall saves into its main database on a Windows laptop. “The database is unencrypted. It’s all plain text,” Hagenah says.⁩ Since Microsoft revealed Recall in mid-May, security researchers have repeatedly compared it to spyware or stalkerware that can track everything you do on your device. “It’s a Trojan 2.0 really, built in,” Hagenah says, adding that he built TotalRecall—which he’s releasing on GitHub—in order to show what is possible and to encourage Microsoft to make changes before Recall fully launches.

The company unveiled Recall as part of a Surface laptop event last month. The tool continuously takes screenshots of whatever’s happening on your PC. Recall is intended to allow people to “retrieve” things you’ve done on your machine—whether it’s web pages you’ve visited or messages you’ve been sent—using natural language search queries. Microsoft’s description of the tool says Recall could be used to search for recipes you’ve looked at online but whose websites you’ve forgotten.

TotalRecall, Hagenah says, can automatically work out where the Recall database is on a laptop and then make a copy of the file, parsing all the data as it does so. While Microsoft’s new Copilot+ PCs aren’t out yet, it’s possible to use Recall by emulating a version of the devices. “It does everything automatically,” he says. The system can set a date range for extracting the data—for instance, pulling information from only one specific week or day. Pulling one day of screenshots from Recall, which stores its information in an SQLite database, took two seconds at most, Hagenah⁩ says.

Included in what the database captures are screenshots of whatever is on your desktop—a potential gold mine for criminal hackers or domestic abusers who may physically access their victim’s device. Images include captures of messages sent on encrypted messaging apps Signal and WhatsApp, and remain in the captures regardless of whether disappearing messages are turned on in the apps. There are records of websites visited and every bit of text displayed on the PC. Once TotalRecall has been deployed, it will generate a summary about the data; it is also possible to search for specific terms in the database.

Hagenah⁩ says an attacker could get a huge amount of information about their target, including insights into their emails, personal conversations, and any sensitive information that’s captured by Recall.

Hagenah’s work builds on findings from cybersecurity researcher Kevin Beaumont, who has detailed how much information Recall captures and how easy it can be to extract it. Beaumont also says he has built a website where a Recall database can be uploaded and instantly searched. He says he hasn’t released the site yet, to allow Microsoft time to potentially change the system. “InfoStealer trojans, which automatically steal usernames and passwords, are a major problem for well over a decade—now these can just be easily modified to support Recall,” Beaumont writes.

The criticisms come as hacks of Microsoft systems have led to various US government data breaches; Nadella has said security should be Microsoft’s “top priority.” Microsoft did not respond to WIRED’s request for comment about the security features of Recall by the time of publication.

Recall’s privacy pages say it is possible to disable saving screenshots (effectively turning Recall off), pause the system temporarily, filter applications where screenshots are taken, and delete what is gathered at any time. Recall runs on the laptop itself, storing data it captures on the device and not sending this information to Microsoft’s servers. Hagenah⁩ says this claim appears to be true, with no signs that data is sent to Microsoft.

Microsoft is, at least, aware of some of the possible privacy and security-related issues with Recall: Its help pages say the system does not perform any content moderation on what is contained in the images it saves. This means, Microsoft says in the guide, that it won’t “hide information such as passwords or financial account numbers.” Security researchers have already been able to extract passwords from Recall.

Recall’s main database is stored on the laptop’s system directory, and while it needs administrator rights to access, privilege escalation attacks have been around for years, making it theoretically possible for an attacker to gain initial access to a device remotely.

Hagenah⁩ says that in cases of employers with “bring your own devices” policies, there’s a risk of someone leaving with huge volumes of company data saved on their laptops. That’s a particular risk if they’re disgruntled or leave on bad terms, he says. The UK’s data protection regulator, the Information Commissioner’s Office, has asked Microsoft to provide more details about Recall and its privacy.

While Recall remains as a “preview” feature and, according to Microsoft’s small print, could change before it launches, Beaumont writes in his research that the company “should recall Recall and rework it to be the feature it deserves to be, delivered at a later date.” He adds: “They also need to review the internal decisionmaking that led to this situation, as this kind of thing should not happen.”

This Hacker Tool Extracts All the Data Collected by Windows’ New Recall AI

Generative AI tools such as OpenAI’s ChatGPT and Microsoft’s Copilot are becoming part of everyday business life. But they come with privacy and security considerations you should know about.

Generative AI tools such as OpenAI’s ChatGPT and Microsoft’s Copilot are rapidly evolving, fueling concerns that the technology could open the door to multiple privacy and security issues, particularly in the workplace.

In May, privacy campaigners dubbed Microsoft’s new Recall tool a potential “privacy nightmare” due to its ability to take screenshots of your laptop every few seconds. The feature has caught the attention of UK regulator the Information Commissioner’s Office, which is asking Microsoft to reveal more about the safety of the product launching soon in its Copilot+ PCs.

Concerns are also mounting over OpenAI’s ChatGPT, which has demonstrated screenshotting abilities in its soon-to-launch macOS app that privacy experts say could result in the capture of sensitive data.

The US House of Representatives has banned the use of Microsoft’s Copilot among staff members after it was deemed by the Office of Cybersecurity to be a risk to users due to “the threat of leaking House data to non-House approved cloud services.”

Meanwhile, market analyst Gartner has cautioned that “using Copilot for Microsoft 365 exposes the risks of sensitive data and content exposure internally and externally.” And last month, Google was forced to make adjustments to its new search feature, AI Overviews, after screenshots of bizarre and misleading answers to queries went viral.

**Overexposed**

For those using generative AI at work, one of the biggest challenges is the risk of inadvertently exposing sensitive data. Most generative AI systems are “essentially big sponges,” says Camden Woollven, group head of AI at risk management firm GRC International Group. “They soak up huge amounts of information from the internet to train their language models.”

AI companies are “hungry for data to train their models,” and are “seemingly making it behaviorally attractive” to do so, says Steve Elcock, CEO and founder at software firm Elementsuite. This vast amount of data collection means there’s the potential for sensitive information to be put “into somebody else’s ecosystem,” says Jeff Watkins, chief product and technology officer at digital consultancy xDesign. “It could also later be extracted through clever prompting.”

At the same time, there’s the threat of AI systems themselves being targeted by hackers. “Theoretically, if an attacker managed to gain access to the large language model (LLM) that powers a company's AI tools, they could siphon off sensitive data, plant false or misleading outputs, or use the AI to spread malware,” says Woollven.

Consumer-grade AI tools can create obvious risks. However, an increasing number of potential issues are arising with “proprietary” AI offerings broadly deemed safe for work such as Microsoft Copilot, says Phil Robinson, principal consultant at security consultancy Prism Infosec.

“This could theoretically be used to look at sensitive data if access privileges have not been locked down. We could see employees asking to see pay scales, M&A activity, or documents containing credentials, which could then be leaked or sold.”

Another concern centers around AI tools that could be used to monitor staff, potentially infringing their privacy. Microsoft’s Recall feature states that “your snapshots are yours; they stay locally on your PC” and “you are always in control with privacy you can trust.”

Yet “it doesn’t seem very long before this technology could be used for monitoring employees,” says Elcock.

**Self-Censorship**

Generative AI does pose several potential risks, but there are steps businesses and individual employees can take to improve privacy and security. First, do not put confidential information into a prompt for a publicly available tool such as ChatGPT or Google’s Gemini, says Lisa Avvocato, vice president of marketing and community at data firm Sama.

When crafting a prompt, be generic to avoid sharing too much. “Ask, ‘Write a proposal template for budget expenditure,’ not ‘Here is my budget, write a proposal for expenditure on a sensitive project,’” she says. “Use AI as your first draft, then layer in the sensitive information you need to include.”

If you use it for research, avoid issues such as those seen with Google’s AI Overviews by validating what it provides, says Avvocato. “Ask it to provide references and links to its sources. If you ask AI to write code, you still need to review it, rather than assuming it’s good to go.”

Microsoft has itself stated that Copilot needs to be configured correctly and the “least privilege”—the concept that users should only have access to the information they need—should be applied. This is “a crucial point,” says Prism Infosec’s Robinson. “Organizations must lay the groundwork for these systems and not just trust the technology and assume everything will be OK.”

It’s also worth noting that ChatGPT uses the data you share to train its models, unless you turn it off in the settings or use the enterprise version.

**List of Assurances**

The firms integrating generative AI into their products say they’re doing everything they can to protect security and privacy. Microsoft is keen to outline security and privacy considerations in its Recall product and the ability to control the feature in **Settings > Privacy & security > Recall & snapshots**.

Google says generative AI in Workspace “does not change our foundational privacy protections for giving users choice and control over their data,” and stipulates that information is not used for advertising.

OpenAI reiterates how it maintains security and privacy in its products, while enterprise versions are available with extra controls. “We want our AI models to learn about the world, not private individuals—and we take steps to protect people’s data and privacy,” an OpenAI spokesperson tells WIRED.

OpenAI says it offers ways to control how data is used, including self-service tools to access, export, and delete personal information, as well as the ability to opt out of use of content to improve its models. ChatGPT Team, ChatGPT Enterprise, and its API are not trained on data or conversations, and its models don’t learn from usage by default, according to the company.

Either way, it looks like your AI coworker is here to stay. As these systems become more sophisticated and omnipresent in the workplace, the risks are only going to intensify, says Woollven. “We're already seeing the emergence of multimodal AI such as GPT-4o that can analyze and generate images, audio, and video. So now it's not just text-based data that companies need to worry about safeguarding.”

With this in mind, people—and businesses—need to get in the mindset of treating AI like any other third-party service, says Woollven. “Don't share anything you wouldn't want publicly broadcasted.”

AI Is Your Coworker Now. Can You Trust It?

When a drug kingpin named Microsoft tried to seize control of an encrypted phone company for criminals, he was playing right into its real owners’ hands.

Inside the Biggest FBI Sting Operation in History

YouTube remains the only major US-based social media platform available in Russia. It’s become "indispensable" to everyday people, making a ban tricky. Journalists and dissidents are taking advantage.

In December 2020, Vladimir Putin held his end-of-the-year press conference as normal. Despite the Covid-19 pandemic still raging in Russia and worldwide, the president insisted that things would be OK. He had set aside 350 billion rubles ($4.8 billion) “to give social benefits to people, families, both children, doctors, and students.”

Two weeks later and nearly 1,000 miles away, the opposition Anti-Corruption Foundation uploaded a video to its YouTube channel. In it, Alexei Navalny introduces the world to “Putin’s palace,” which he dubbed “the world’s largest bribe.”

Over nearly two hours, Navalny walks viewers through the trappings of the palatial datcha, on the Black Sea coast. A 2-million-ruble sofa, a 2-million-ruble vanity, a dining table worth nearly 4 million rubles. The ubiquitous gold leafing on the walls, a private theater, a velvet-lined hookah bar, a casino, even a _Dance Dance Revolution_ room. The construction of such a lavish abode would have cost Putin and his oligarchic backers over 100 billion rubles ($1 billion) of money pilfered from the Russian state, Navalny tells the viewer. Money that could have gone to people, families, children, doctors, and students.

The Russian-language video exploded, far surpassing anything Navalany and the Foundation had ever produced. Within weeks, it blew past 100 million views, and it has added another 32 million views since then—nearing the population of Russia itself.

The Foundation’s popularity on the world’s largest video-sharing platform had been growing massively over the preceding year—particularly as Russians were stuck indoors, and as the Kremlin intensified its crackdown on the independent press.

“Twenty years of total government control have turned television into an incessant marathon of disgraceful propaganda, lies, and censorship,” as Navalny told his viewers in a 2020 video. Behind him, a stack of TVs broadcast different clips of Russian TV presenter Vladimir Solovyov, a ubiquitous face on the Russia1 channel.

In December of that year, Navalny began his regular video dispatches with a startling introduction: “Hi, it’s Navalny. I know who wanted to kill me. I know where they live. I know where they work. I know their real names.”

Working with investigative journalism outfit Bellingcat, Navalny had posed as a Kremlin official in order to coax a confession out of one of his would-be assassins. The alleged agent recounted how the team of Russian operatives applied the nerve agent Novichok to Navalny’s underwear. Navalny recorded the whole thing and posted it directly to his YouTube channel.

In a follow-up video, Navalny continues to break down the failed murder plot in front of a wall featuring each of the men’s photos, maps of the operation, and red twine connecting everything.

In the West, Navalny’s appeal was often shortened to that of simply an opposition figure—a divisive one, for some, who became the de facto alternative to Putin after the other alternatives were killed or jailed.

But in the years leading up to his 2021 imprisonment and death in February 2024, Navalny became much more than just a politician—and the media apparatus he became the face of became bigger than him.

While Russia has grown considerably more autocratic in recent years, with its elections even more tightly scripted and its media ever more controlled by the Kremlin, YouTube remains one of the few US-based social media sites still available in the country and a reliable and effective conduit to the Russian public. It could be key to determining what happens next in Moscow—if YouTubers can stay ahead of Kremlin censorship.

“With minimal resources, we were essentially able to establish a direct competitor of Putin’s propaganda television,” says Vladimir Milov. “We are still behind them, but we are breathing down their back.”

Milov is a former deputy minister with the Russian government, a longtime Navalny associate, and a YouTube creator in his own right. I caught up with him in Vilnius, where he’s been exiled since Moscow began criminal proceedings against him. Last year, Milov was convicted by the Russian courts of disseminating “war fakes” and sentenced to eight years in prison in absentia.

Milov’s regular updates, which tend to focus on economics and finance, go out to his 500,000-plus subscribers, and often rack up around a million views. Leonid Volkov, another Navalny associate who helps run the Anti-Corruption Foundation, boasts another 165,000 followers. Dasha Navalnaya, Alexei Navalny’s daughter, has her own channel where she interviews GenZ Russians about the state of their country.

Above that, independent media have found a second home on YouTube as well. Meduza, an independent media organization that had to decamp to Latvia after being declared a “foreign agent” under Russian law, has nearly 800,000 subscribers. The BBC’s Russian service, which left the country amid a 2022 media crackdown, broadcasts regularly to its 2.5 million subscribers.

Cobbling together the analytics across these various channels, Milov says these opposition YouTubers have about 10 million to 15 million dedicated viewers inside Russia—while another 20 million viewers may stumble across their content from time to time.

​”Our message gets through to the Russian people,” Milov says. “We are serving as a very effective alternative broadcasting tool that really competes with Putin's propaganda.”

VCIOM, a pollster loyal to the Kremlin, has found that a plurality of Russians still prefer to get their news from heavily censored Russian television, but their trust in the format has plummeted, from nearly 50 percent in 2016 to just 25 percent today. Meanwhile, trust in online news has jumped by roughly the same degree.

Meanwhile, as Western programming becomes rarer and rarer, many Russians are turning to YouTube for their primary source of entertainment—especially for kids’ programming. Eight of the 10 most-subscribed Russian YouTube channels, according to rankings compiled by analytics firm Socialblade, are targeted at young children.

So Russians are losing faith in the state-approved TV news, increasingly trusting of online media, and are looking to YouTube for entertainment. It has delivered tens of millions of people right at the opposition’s doorstep.

Even if most Russian YouTube users are looking for content to pacify their children, bringing them on the platform has enormous benefits, Milov says. “There's this magical black box which is called YouTube recommendations.” He says plenty of his followers don’t even watch his videos—they listen. If Russians are letting the algorithm choose their next video while they cook dinner or fold laundry, it increases the chances his voice will reach a Russian who has never heard him before.

Milov stresses that YouTube isn’t just a one-way service: Because it allows users to comment and chat anonymously, it provides an extraordinary chance for regular Russians to express themselves without fear of censorship.

“The amount of our feedback is enormous,” he says. “Just myself, alone, I literally get messages, every day, from at least hundreds of people from across the country. When something serious happens? Thousands.” Sometimes, Milov says, his first indication that something terrible has happened in Russia is seeing just how many unread messages he has in his YouTube inbox.

Milov says this feedback reinforces the idea, supported even by Kremlin-approved pollsters, that opposition to the war in Ukraine is growing. But it also provides some important details and nuance. “So this is like, I would say, an enormous focus group, with which you can also communicate. You can ask them questions back.” He chuckles, thinking of the notorious Russian security and intelligence agency: “You know, the FSB would kill for this kind of information.”

“Obviously, the question is, why didn’t Putin shut down YouTube?” Milov says. “It’s easier said than done.”

In recent years, Moscow has deployed an array of strategies to cow and kill independent media and the open internet in Russia. Platforms like Facebook, Twitter, and TikTok have been blocked altogether. Independent media like Meduza, TV Rain, and The Insider have been declared “undesirable” or labeled “foreign agents.”

Through it all, YouTube has survived.

Milov says the Kremlin was too slow to move on YouTube. By the time Moscow was banning other popular Western platforms, the Google-owned video platform had become indispensable to everyday Russians. “They kind of let the genie out of the bottle,” Milov says.

“YouTube is mommies showing cartoons to kids, teenagers are watching music videos, people are watching comedians, elderly folks watching old Soviet movies, which are widely available there, and so on,” he says. “And you shut it all down? So you have these empty evenings now, from this point on.”

Unable to disrupt YouTube, the Kremlin tried desperately to compete with it.

Moscow had high hopes for Rutube, a long-suffering YouTube clone which was relaunched in 2020 after a merger with the media arm of state-controlled energy giant Gazprom. If the site’s “top videos” section is to be believed, it hasn’t worked—some had racked up view counts in the mid-thousands.

VK, Russia’s answer to Facebook, has fared slightly better with its video-sharing platform, and it is rife with pro-Kremlin broadcasters. But even its most popular channels have just a tiny fraction of the biggest Russian-language YouTube accounts.

“It's like a big room, but it's empty,” Milov says of these Kremlin-backed alternatives.

Having failed to compete with his online critics, Milov believes Putin opted for a more direct strategy. Just days before I arrived in Vilnius, thugs appeared outside the home of Leonid Volkov, former chair of the Anti-Corruption Foundation and Nalvany chief of staff. Armed with hammers, they savagely beat him. Lithuanian intelligence believe the men arrested were operating on orders from Russia. A week after the attack, Volkov was back on YouTube, his arm in a sling, “I am not going to stop—although I will gesticulate less in the coming weeks,” he said.

Milov emailed me the day I arrived to apologize that, due to his new security protocols, he wouldn’t be able to show me around his home broadcast studio.

Last November, the YouTube channel Volgograd Watch uploaded a new video trying to answer a difficult question: “When will the mobilized be returned?”

Subscribers of the channel were asking when Russia’s conscripts on the front line might return home. Evgeny Kochegin, host of the channel, responded by providing a litany of links recommending anti-war groups and resources on how to dodge the draft. Kochegin himself fled conscription and was convicted in absentia for disseminating “fake news.”

In exile, Kochegin has tried to help others avoid being pressed into military service, and took to YouTube to do it, racking up a modest 30,000 subscribers. In May, however, Kochegin found YouTube to be an unreliable partner. His video was blocked for Russian viewers.

Agents Media, an independent Russian news outlet, reviewed four different removal notices targeting human rights and anti-war YouTube channels. The news outlet obtained emails from YouTube’s legal team, threatening to take offline an entire channel. In the email to the OVD-Info channel, YouTube wrote that the channel had violated Russian Law #149-FZ and that “if you do not remove the content, Google may be required to block it.”

According to a company transparency report, Google has received hundreds of thousands of removal requests from the Russian government since 2022, including nearly 67,000 on the grounds of “national security,” over 200 for “government criticism,” and one for violating the electoral law.

More than 1,000 applications from Moscow specifically cited Law #149-FZ, asking for the removal of nearly 1 million pieces of content—which could include channels, videos, or comments. Google reports that it approved more than 80 percent of those removal requests.

An open letter published in May, signed by major Russian NGOs and Reporters Without Borders, called on Google to stop adhering to these requests from Moscow. “We are very concerned that the company appears to be helping the Russian censor to silence human rights and anti-war voices,” they wrote.

In a statement, a spokesperson for YouTube insisted that if the company has concerns that legal requests are being used to silence dissidents, “we will push back”—noting that Russian courts have fined Google in the past for refusing to comply with its orders.

YouTube says it has removed more than 12,000 channels and over 140,000 videos relating to the war in Ukraine for violating the platform’s policies—including pro-Putin propagandists, Kremlin-controlled media outlets, and content from the Russian Ministry of Foreign Affairs.

The YouTube spokesperson, however, ignored questions as to why it has been so willing to respond to requests under Law #149-FX.

Whether their channels are at risk from legal requests from Moscow, or whether Moscow blocks their channels entirely—the Russian dissidents are always making contingency plans. “They will not kill us by just shutting down one thing and making the other cooperate with the regime,” Milov says. “We'll navigate through.”

One such plan is to wind up right where the problem began: on TV.

Not long after Russia invaded Ukraine in February 2022, a group calling itself the Denis Diderot Committee began calling on major satellite television providers to stop beaming Kremlin propaganda into Russia and Europe — and replace it with real news.

“That's the ultimate goal of our effort—to actually provide alternative media channels into the Russian television space that are not controlled by the Russian government,” Jim Phillipoff, one half of the committee, told WIRED at the time.

Two years on, the committee scored a huge win. Satellite provider Eutelsat took up the committee’s challenge and inked a deal with Reporters Without Borders to broadcast 11 channels of independent Russian news and content into the region, with the option to add 14 more. They call it the Svoboda Satellite Package—using the Russian word for “freedom.”

Phillipoff told WIRED this week that they have begun broadcasting into 4.5 million households in Russia, and another 61 million households in the broader region—and they hope to grow that number by securing space on other regional satellites.

While the satellite is simply rebroadcasting some existing channels, Philipoff says they’ve also created some channels from scratch—in some cases made up of dissident Russians’ YouTube content. They’ve been in talks with Volkov to help rebroadcast some of the work from the Anti-Corruption Foundation and other opposition groups. (They have also explored how to cooperate with eQsat, a like-minded effort to broadcast digital news files using satellites, which WIRED revealed last September.)

While Russian television remains under the thumb of the Kremlin, and Google remains willing to block content at its behest, Svoboda exists to reject the censorship altogether, Phillipoff says.

“The goal of our project is to circumvent this control.”

Source

Wired