This is the web version of Data Sheet, a daily tech newsletter. Sign up to have it delivered to your inbox for free.
When we hear about cybercrime, we usually hear from the leading players, companies like Crowdstrike who nailed the Russians for stealing DNC emails in 2016. Or Microsoft warning that the Russians were trying to hack the 2018 election campaigns. Or FireEye, which revealed last month that it itself had been invaded by nation-state hackers (who turned out to be Russians).
But as we learn from that latest incident, we cannot guarantee cyber security by relying solely on the big names.
FireEye had uncovered the tip of what is now believed to be the largest and most harmful hack in cybersecurity history, a hack that compromised the computer networks of hundreds of major corporations and government agencies, including the U.S. Treasury Department, the State Department and the Department of Homeland Security. The attack is called SolarWinds, after an obscure software developer in Austin, Texas, who was the starting point of the entire disaster.
As Data Sheet’s own Robert Hackett and our tech colleague David Z. Morris explain in their new article on the SolarWinds attack, Russian hackers were able to penetrate so many networks simply by putting a back door into security software the company produced and to its many customers. in the entire country.
Their deep dive explains not only how it happened, but why. In particular, note David and Robert, the SolarWinds hackers didn’t go for the usual credit card numbers and email addresses that most cyber thieves are looking for. Instead, the hackers went for a lot of high-quality internal information: emails containing company and government secrets, the source code that underpins Microsoft software, and the like.
The attack also undermines not only the reliance on one company, SolarWinds, but arguably the entire cybersecurity structure in the United States, with its patchwork of government agencies, well-known security companies, thousands of smaller third-party vendors, and internal IT security. efforts.
“Most industry experts see the decentralized, market-driven structure of US cybersecurity as a source of flexibility and innovation,” write David and Robert. “But in the SolarWinds debacle, they also see the weaknesses of the system in full screen. In this mega breach, lack of financial incentives, lack of transparency, underinvestment in training, and outdated cost savings each played a role.”
We are all familiar with the science fiction trope of a computer getting so smart that it takes on a mind of its own. That fantasy feels all too realistic these days, thanks to advances in Natural Language Processing (NLP). In this week’s Brainstorm podcast, hosts Michal Lev-Ram and Brian O’Keefe explore what it means to teach a computer to understand and even ‘think’ like a human. What are the innovative possibilities that this creates? What are the dangers? Listen to the episode here.