Friday, November 22, 2019


An audit sampling tool, not a security guarantee.
CISA Announces Open Source Post-Election Auditing Tool
Called Arlo, the newly introduced auditing tool is being created by non-partisan, non-profit firm VotingWorks, which is committed to developing secure election technology. The tool includes support for various types of post-election audits across numerous voting systems, including major vendors.
Arlo is built on auditing work performed by Colorado, which in 2017 implemented rigorous Risk-Limiting Audits (RLAs) and conducted audits in over 50 of its 64 counties.
The Arlo open source software is being offered for free to state and local election officials, and their private sector partners. State and local jurisdictions can also use a hosted Software-as-a-Service version of the tool, for a fee.
Arlo provides an easy way to perform the calculations needed for the audit: determining how many ballots to audit, randomly selecting which ballots will be audited, comparing audited votes to tabulated votes, and knowing when the audit is complete,” CISA says.




Another pendulum swing!
PA Supreme Court – Police Can’t Force You to Tell Them Your Password
EFF: “The Pennsylvania Supreme Court issued a forceful opinion today holding that the Fifth Amendment to the U.S. Constitution protects individuals from being forced to disclose the passcode to their devices to the police. In a 4-3 decision in Commonwealth v. Davis, the court found that disclosing a password is “testimony” protected by the Fifth Amendment’s privilege against self-incrimination.
EFF filed an amicus brief in Davis, and we were gratified that the court’s opinion closely parallels our arguments. The Fifth Amendment privilege prohibits the government from coercing a confession or forcing a suspect to lead police to incriminating evidence. We argue that unlocking and decrypting a smartphone or computer is the modern equivalent of these forms of self-incrimination…”




What happens when “double secret probation” goes public?
California Stepping Up Its Probe Into Facebook Privacy Practices
In a new lawsuit, California Attorney General Xavier Becerra is accusing Facebook of failing to adequately comply with information and subpoena requests related to the company’s privacy practices. The investigation into Facebook privacy practices has been ongoing since 2018, but had never before been publicly disclosed until now. The state attorney general’s office had been working on a private investigation, but had been repeatedly stonewalled by Facebook for key documents and information, and had no other recourse than to make the investigation public.




Is there a polite way to surveil you employees and customers? Just an extension of “This call may be monitored for training purposes?”
Uber Embraces Videotaping Rides, Raising Privacy Concerns
The New York Times – “The company says it is increasing the recording of rides to settle disputes between drivers and passengers and to improve safety… Uber began the video recording program in Texas in July, and is conducting smaller tests of the program in Florida and Tennessee. In November, it announced a similar effort in Brazil and Mexico to allow riders and drivers to record audio during a trip. The audio recording feature, first reported by Reuters, is managed by Uber, and begins a recording if either the rider or driver requests it. At the end of the trip, the rider or driver has the option to send the recording to Uber for review, but cannot save it themselves, a safeguard Uber built to prevent riders and drivers from recording each other and posting the clips online, the company said. Uber’s video recording feature is a partnership with Nauto, a technology company that uses artificial intelligence to analyze video from vehicles…”




Interesting ways to look at AI as it evolves.
The Role of AI in the Future of Business Intelligence
Today’s businesses leverage the power of AI in many ways, from call centers deploying AI-based chatbots to banks using deep learning to analyze countless data points in seconds and detect fraud.
AI also has the potential to change the dynamics of analytics. Conventional data analytics focused on descriptive analytics or analyzing data to report what happened. The present generation of AI-enabled analytics tools enable predictive analytics or using data to decipher future insights. This, however, is based on “best guesses” with behavioral and historical data used to guess probabilities.
Prescriptive analytics is all set to take over in the near future. AI-powered prescriptive analytics tools would scour through vast swathes of data and enable users to prescribe various possible actions and advise viable solutions. Prescriptive analytics not just predicts, but offers sound advice as well, and explains why things will happen the way it will or does.



No comments: