Skip to content Skip to navigation

Friday Cyber News, February 5 2016

Cyber technology-related news and links from around the web, for the week of 1/30 - 2/5:

1. The internet is learning what you like about it: AI in the form of deep learning is helping with 15% of Google searches through the RankBrain system, and is growing more adept at handling novel queries than Google's rule-based system. Deep learning algorithms may be the only way to rout fake business listings from Google's maps and search listings: fake locksmiths have proliferated online, supported by photoshopped storefronts and Israeli recruitment, and human (Google employee) intervention can't keep up. [Wired; NYTimes]

2. Football players are some of the most intensively tracked employees in the country: their uniforms include small sensors that record their position, velocity, and other data as they move around the field. Those sensors may be coming to an office near you soon, as useful patterns emerge from the monitoring of truck drivers, Wall Street traders, and Hitachi engineers. [Fusion]

3. The University of California system moved to implement a data-tracking program across its networks to address cybersecurity concerns. Researchers, many of them computer scientists, have protested the program, due to a lack of transparency over how the program collects and stores data, and whether certain research activities would be curtailed. [NYTimes]

4. China may have an incentive to stop North Korea's hackers, who are often able to operate within China's borders. Doing so would help South Korea, as well as reinforce China's message of cyberspace sovereignty. [CS Monitor]   

5. The Berkman Center's Cybersecurity Project discusses the "going dark" debate, with contributions from academia and US intelligence, coming to the conclusion that end-to-end encryption is unlikely to be universally adopted, and metadata and the proliferation of "channels" (via IoT devices) will provide plenty of unencrypted data for those who want to surveil. Even some apps that promise end-to-end encryption and anonymity aren't delivering, UCL researchers have found, so "going dark" concerns are still theoretical. [Harvard;]

6. The NSA is planning a reorganization, bringing Information Assurance and Signals Intelligence functions together under one operations directorate. Bringing together "spying" and "defending against spying" functions should help the agency distribute information internally, but some caution that tech companies will be less likely to share information knowing that it will be used for both purposes. [Washington Post]

7. The European Commission has proposed rules for regulating virtual currency platforms, including national supervision to prevent their use for money laundering or to fund terrorism. [Infosecurity Magazine]

8. AI bots are active on online dating and gaming websites, scamming users into buying paid features and credits. [Rolling Stone]

9. The successor to Safe Harbor has a reassuring name: Privacy Shield. But will it address Europeans' legal concerns while stopping short of offering them the same protections against government snooping as US citizens' data? [CS Monitor]

10. In a move that may rehabilitate its reputation, GCHQ would like to help with the spam email problem, partially because spam emails interfere with their SIGINT efficiency. [Vice]

In the Bay Area? Join us Feb. 8th at 5pm for a seminar on behavioral research and decision-making in the age of the internet: proprietary datasets, algorithmic bias, and social media experimentation will be discussed by panelists Ashish Goel (Management Science & Engineering), Rob MacCoun (Law), Jeff Hancock (Communication), and Naomi Grewal (SurveyMonkey). In SLS 180 (Law School classroom building)


Stanford Cyber Initiative

(To suggest an item for this list, subscribe, or unsubscribe, please email This email contains links that may not work if HTML is not supported in your mailbox. You can also subscribe or view news from past weeks at