FFT news digest — 19 May 2023
FFT news digest — 19 May 2023
Ransomware
This year has seen a sharp increase in the number of ransomware attacks, with the data from 452 new victims appearing on leak sites in March alone. That represents a 60% increase on the previous year, according to Corvus Insurance . However, despite the rise, Corvus adds that many organisations are better prepared to defend themselves against attacks. In other words, assuming everything will be OK is a good way to become a victim. One trend highlighted by Coveware’s Quarterly Ransomware report is a focus on bigger payouts, or “big game hunting” as they put it. Phishing continues to be the primary threat - and it appears to be mainly directed at medium and large organisations. Like Corvus, Coveware says it’s seen more of them able to deflect the most damaging aspects of attacks successfully. Meanwhile, the UK’s National Cyber Security Centre and the country’s data protection regulator have joined forces to try to persuade ransomware victims not to hide their misfortune. “If attacks pass by without full investigation and information sharing, particularly with those who can help mitigate it, everything definitely won’t be OK. Keeping your cyber incident a secret doesn’t help anyone except the criminals,” they wrote in a joint article.
Threats
Phishing : One of the most comprehensive surveys of phishing activity shows a relentless upwards trend with almost double the number of email subjects compared to a year ago. The Anti-Phishing Working Group says the growth results from a mix of new entrants, improvements in phishing kits, and a rise in the number of successful attacks. Invoice : Armorblox has a good example of the growing sophistication of attacks. In this case, a fake invoice is followed up with a fake email from an executive insisting that payment should be processed immediately. Super Mario : We should all know by now that downloading pirated movies and software is not only illegal but a huge security risk. As the new Super Mario Bros movie illustrates. ReasonLabs PDF Toolbox : We should be cautious even with widely used browser extensions. PDF Toolbox for Chrome has more than two million users but a researcher has just discovered apparently malicious code inside it. Almost Secure Email security : Attackers are impersonating email security providers in a bid to lure recipients into revealing their user credentials. Cofense LinkedIn : A survey from NordLayer suggests 52% of US-based businesses have already experienced a LinkedIn scam this year. LinkedIn has a safety guide that is well worth reading. macOS : SentinelOne spotted a new threat to Apple Macs. It’s a version of a sophisticated tool that has been used against Windows devices for several years. The Mac versions arrived as a fake PDF document and a remote support tool. Teams : Scumbags have found new ways to exploit Microsoft Teams and steal user credentials. The latest tactic abuses the Tabs feature to display phishing pages. (Tabs is used to synchronise items between Teams and Calendar.) Proofpoint
AI
Artificial Intelligence assistants may be able to change our views without us realising what’s going on, according to a new study . One researcher described this as “latent persuasion” and the only defence is awareness about it. In other AI news; Senate : Top lines from a Senate hearing on AI which featured OpenAI’s CEO. Government regulation will be “critical,” particularly around elections, Sam Altman said. He added that he is worried the AI industry could cause “significant harm to the world.” Update : OpenAI’s ChatGPT subscription version is getting a bunch of new features including a direct connection to the internet so that it can act as an online research assistant. Assistant : US researcher, Daniel Messier, says AI assistants will know everything about us and become our primary interface with the world. Open-source : As the technology giants battle for AI supremacy, Meta has decided to make public the engine that powers its AI platform. Google and others say it’s a risky move. New York Times Accessibility : iPhones will soon be able to speak in their owners’ voices after just 15 minutes of training. It’s seen as a major advance for sufferers of degenerative diseases like ALS. Rogue : A social media influencer created an AI version of herself. It has now gone rogue and begun to engage in explicit conversations for $1 a minute. PetaPixel
In brief
Pentagon : Some salutary lessons from the case of Jack Texeira, the US airman accused of leaking classified information. It’s emerged that he was warned twice after his superiors caught him up to no good (including taking notes and putting them in his pocket). It’s often the case that theft of information is part of a pattern that turns out to have been well-known, so it’s obviously essential not to ignore the warning signs. Philadelphia Inquirer : The newspaper says a cyberattack caused the biggest disruption to its operations in 27 years. Zip : Just what the world needs; a new top level domain (i.e. like .com) called zip. Silent Push Labs says it already spotted what appears to be a phishing page at microsoft-office[.]zip. Twitter : The sister of an activist who disappeared in 2018 has filed a lawsuit against Twitter and the Saudi government alleging the social media network allowed the kingdom to target dissidents including her brother by failing to protect confidential data. A former Twitter employee was convicted last year of spying for Saudi Arabia. Privacy : What privacy? An anonymous complainant is suing Google saying it unlawfully collects health data, including searches related to abortion. And the US Federal Trade Commission says a popular fertility app gave advertisers pregnancy data without permission. JDSupra has a round up of related cases. Surveillance : Two stories about the extent of government surveillance. Wired reports on the expansion of a UK government programme to collect people’s internet records. And Motherboard explains how the US is using an AI tool to analyse the social media activity of US citizens and refugees. TikTok : Montana became the first US state to ban the Chinese-owned app in what is likely to be a test of technical expertise - and the extent of the country’s First Amendment. The ban is due to come into force from 2024. Toolkit : The Global Cyber Alliance’s cybersecurity toolkit is now available through a consortium of university-based clinics. It’s designed to provide cybersecurity assistance to under-resourced organisations while giving students hands-on experience. It’s mainly US-focussed but expanding internationally. Double-screen : Want a free telly? If you’re in the US, it’s yours as long you agree to let it track everything you watch and also install a second screen designed to show you constant adverts. The Verge