|
|
|
|
FICO Launches Behavioral Biometrics; a New Unicorn; Automation Altering HR
|
|
|
|
|
|
|
|
|
Liz Lasher, vice president of fraud and financial crimes at FICO. PHOTO: FAIR ISAAC CORP.
|
|
|
|
FICO rolls out behavioral biometrics. The creator of the widely used FICO credit score, Fair Isaac Corp., is set to unveil new AI tools to make it easier for banks to verify customers and reduce fraud, reports WSJ Pro’s Jared Council. One of the tools uses behavioral biometric authentication. The behavioral component can track patterns such as how fast a person types or the mobile device and language setting used for logging in. Machine-learning algorithms recognize patterns over time and can detect if something looks awry.
The new FICO tools take widely multifactor authentication, where users are prompted to enter an additional piece of information, to the next level. Only 30% of people use multifactor authentication to log into accounts, according to Chubb Ltd. Thus the need for more frictionless authentication. “Right now it’s an everything-mobile world,” Liz Lasher, vice president of fraud and financial crimes at FICO. “So when you think about the need to digitally onboard customers seamlessly without any face-to-face verification, there’s an element of being able to prove, secure and establish a digital identity. And this technology allows us to do that.”
|
|
|
|
|
|
|
|
Tel Aviv, where Riskified Ltd. got its start, on a sunny day in September. PHOTO: KOBI WOLF/BLOOMBERG NEWS
|
|
|
|
A new AI unicorn. Tel Aviv data-security company Riskified Ltd. received $165 million in a funding round that pushed its valuation above $1 billion, reports the Journal’s Ted Bunker. Private-equity firm General Atlantic led the round. Merchants use Riskified's software to secure their payment systems and guard against e-commerce scams. E-commerce fraud is expected to cost retailers $6.4 billion annually by 2021. A bigger hit comes from false declines—where valid transactions are incorrectly tagged as fraudulent. That’s expected to cost retailers $443 billion a year in lost business.
|
|
|
|
|
|
|
|
|
“Facial recognition has important roles—for example, finding lost or displaced children. There are use cases, but they need to be underpinned by values.”
|
|
— Michael Kratsios, the Chief Technology Officer of the United States, during an interview last week at Stanford University, according to IEEE Spectrum.
|
|
|
|
|
|
|
|
|
AI poised to change recruitment and HR. The hiring process is just one example of how human judgment is being augmented by data-driven machines. The subjective task of reading resumes and cover letters and moving favored job candidates through rounds of interviews may be ending. In the future, the FT writes, jobseekers will be assessed by algorithms and exercises that test problem-solving skills, creativity and responses to stress. Catalyte, an AI employment platform conceived by former Clinton White House advisor Michael Rosenbaum, may be an indication of such change. Its database and algorithm used algebra questions to test the skills of Tim Reed, now an automation
engineer at the organization. At first, he thought he failed the test, according to the FT. “He now knows that such questions are part of a hiring algorithm designed not to count correct answers but to measure tenacity, grit and creativity under pressure,” the FT said.
|
|
|
|
|
|
|
Medopad Ltd., a U.K. startup that’s partnering with China’s Tencent Holdings Ltd. on new approaches to Alzheimer’s and diabetes, raised $25 million. (Bloomberg)
Leading researchers admit they don’t have a way to stop adversarial attacks. (VentureBeat)
IBM is calling for what it calls "precision regulation" of facial regulation—limiting potentially harmful uses rather than forbidding use of the technology entirely. (Axios)
Nick Jennings, chair in AI at Imperial College London, joined the advisory board of cyber company Darktrace. (The Information)
Ford Motor hired Scott Griffith, former Zipcar chairman and CEO, to head its driverless vehicle operations. (CNBC)
Mobileye, an automotive sensor company that Intel bought in 2017, is teaming up with Chinese electric car company Nio to develop driverless vehicles for consumers. (TechCrunch)
The U.S. Postal Service is planning to use AI technology to help it process package data 10 times faster. (Reuters)
Booz Allen Hamilton is opening an app-store-like AI marketplace that will include pre-trained AI models that can handle specific tasks, such as recognizing buildings from aerial imagery. (Fortune)
Ecobee, a smart-home thermostat company, unveiled a free suite of machine learning features that can customize temperature settings and cut costs. (VentureBeat)
Promobot, a Russian startup, is marketing robots that buyers can make look like any person in the world. (Fox News)
|
|
|
|
|
|
