Nobody at Google has access to customer’s data, claims the tech giant – Business Insider India
- Olivier Bousquet, head of machine learning at Google, claims that the customer data is ‘siloed and encrypted’, and that no one at Google can access it.
- He points out that there is a need for educating users about AI to ensure that users are able to provide ‘informed’ consent.
- However, he believes that the impetus of regulation AI cannot fall on any individual company. Instead, it’s the shared responsibility of all the stakeholders
Google has been one of the leaders in the field of artificial intelligence (AI). Two years after taking over the reigns of the company from Larry Page in 2015, Sundar Pichai revealed his vision of Google being an AI-first company.
Today, AI is central to Google’s operations. Paving its way in the sector, Google acquired companies like
which has blurred the lines between human and artificial intelligence, demonstrating this with the popular AlphaGo program that
In 2018, Google transformed its research division and
, further indicating just how important AI had become to Google’s future.
However, this change has led to privacy conscious users to question how the company uses its troves of data. The scepticism is understandable given how AI still has a sense of mystery around it and sci-fi movies like the
want of super-AI that could supersede human intelligence.
‘Anything that touches technology is often misunderstood’
“I think maybe the crux of the matter is that, in general, web technologies, not just AI but anything that touches technology, is something that is often misunderstood,” Olivier Bousquet, head of machine learning research at Google told
Reassuring its users and customers at the AI in Action event in Zurich, Google explained how it is working on being responsible when it comes to deploying AI in its products.
One way of doing this is by educating users about AI. “For AI, there is an extra need for education and that’s why we share a lot of the course material that we use internally for people to give consent in a way that they understand what they are giving consent to,” Bousquet added.
He further acknowledges that “there is still a lot more to be done in this area.”
Ethics in AI is not an easily solvable issue
It’s still early days when it comes to AI. As is the case with anything new, it can be hard to determine what is ethical and what isn’t. Google itself has battled with this dilemma for quite some time now – not just in the field of ethics, but technology as a whole.
Most recently, Google had to drop Project Maven after protests and petition signed by thousands of Googlers. The project proposed a contract with the
US Department of Defense
wherein AI would be used for facial recognition purposes.
Bousquet says, “Working with the military for Gmail is one thing. Providing technology that could be used to lead to harm is another thing and that’s where we have to draw the line.”
While it can be hard to understand when and where to draw a line, Bousquet says that “company-wide ethical guidelines” could help as guidance.
Amazon has similarly been under fire for proving facial recognition technology to police forces for monitoring criminal activity.
‘Data collection is not bad per se’
Data is the fuel on which Google is driving its AI journey. Using data to train AI and machine learning models is important is one of the core aspects of these new technologies.
“The fact that we are collecting data is not necessarily bad per se, but, of course, it means that we have a huge responsibility with respect to the privacy of the data and the use of the data,” Bousquet said.
Bousquet reiterates that no one at Google has access to customer data. “When it comes to customers of our cloud, storing their data on our cloud, this data is siloed and encrypted. So that means nobody at Google has access to that data,” he said.
According to him, AI is a shared responsibility saying, “A single company cannot be prescribing how AI should be developed.”See also: After an employee backlash, Google has cancelled its AI ethics board a little more than a week after announcing it