India may soon get a home-grown platform for Artificial Intelligence (AI). This platform will leverage and compile existing databases that are available with the government such as Aadhaar details, land records and soil health among others. Once setup, this platform can be accessed by companies to use AI-based technologies for different sectors such as healthcare and agriculture.

“There will be a platform for AI which will ride on the databases that we have built in. For example, we do have Aadhaar database, we have a land records database, we have a soil health-card database, we have a weather database. And all these data, with regard to say farmers, is linked to geospatial coordinates. And then there is data from land records, the Aadhaar number of farmers and their mobile numbers. So, we will have all this data coming from various disparate data sources available on a common platform, and if this data is made available to AI developers, there is a possibility of building solutions which will guide the farmer to accomplish the best possible agriculture practices,” Abhishek Singh, President and CEO of Digital India (National e-Governance Division (NeGD), Ministry of Electronics and IT (MeitY) and CEO, MyGov, said in an interview to Hindustan Times.

Singh said that the body is working on a framework for the AI platform. He revealed that the platform, that is based on digital health blueprint, is already built and the agency is working on a similar platform for agriculture.

According to Singh, the potential for an AI platform for agriculture is “huge” as it will enable developers design better insurance products and frame intensive agriculture practices.

When asked about privacy concerns, especially around Aadhaar, Singh said the solutions will conform to all the Indian laws and the global standards with regards to privacy. The platform will also ensure that whatever data is available with any authority, whether a public or private, is primarily used for social good and with the consent of the data owners.

“So, whenever we are provisioning usage of anything, including Aadhaar, there will be a consent layer involved. And consent will always be informed, fair and analysed. Whenever a user is going to give his consent for use of Aadhaar details or any other data for any service, he/she will be explained what does giving consent in this case means, what it entails, what else is there and what are the benefits? And after that, if and only if a citizen agrees only then the concerned data will be leveraged to create a framework or solution,” Singh explained.

“In our experience we have seen that wherever there is value offered to a citizen with adequate safeguards for protecting his/her data, citizens by and large do agree to give consent and avail services,” he added while pointing out the data protection bill that has been proposed by the government for users’ security and privacy.

MeitY is gearing for its first official summit on Artificial Intelligence called RAISE 2020, Responsible AI for Social Empowerment 2020, and the summit will be held in New Delhi between April 11 and April 12. Apart from participation key industry people, agencies, and private firms, the event will also host a “Startup Pitchfest”.

“The idea is to bring together the entire community of startups to showcase their products to VCs, to funders, to agencies and to government departments. So, these startups at the Pitchfest can put forward their solutions, there will be buyers available – there will be buyers in the form of VCs, angel investors and there also be government departments. And once they see that there is potential in some applications and the deployment of those solutions they have been working on, either on a commercial scale or in a public environment, agreements and deals will be signed,” Singh said.

The AI solutions showcased by the startups will also be tested by various government departments and further refined, customised to be scaled up. These solutions eventually will be adopted by some state governments or government departments and could also be deployed at a much larger scale nationally.

The post India may soon get a home-grown platform for Artificial Intelligence appeared first on Artificial Intelligence.

Blowing the lid off what could possibly be one of the biggest data breaches of unique identity database of over a billion Indians, a report by Huffpost on Tuesday claimed that UIDAI’ database has been compromised by a software patch that disables critical security features of the software used to enrol new Aadhaar users.

What is Aadhaar software hack?

According to the report, a software patch, which can be bought for as little as Rs 2,500, allows unauthorised persons, based anywhere in the world, to generate Aadhaar numbers.

A patch is a set of code that is used to change a computer program or update, fix, or improve it. This includes fixing bugs. However, it can also be used to introduce vulnerabilities. The report claimed that the Aadhaar software patch allowed users to bypass critical security features such as biometric authentication of enrolment operators to generate unauthorised Aadhaar numbers.

Why is it a security risk?

The report said that Aadhaar patch disables the enrolment software’s GPS security feature which is used to identify the location of enrolment centres. Turning off GPS would allow anyone to use the software to enrol users from anywhere in the world.

How the software patch makes Aadhaar vulnerable to ghost entries?

While the government has trumpeted weeding out of illegal beneficiaries from various central schemes as one of the major successes of Aadhaar, the latest revelations may deliver a body blow to Centre’s tall claims.

The report claims that the patch reduces the sensitivity of the enrolment software’s iris-recognition system, making it easier to spoof the software with a photograph of a registered operator, rather than requiring the operator to be present in person.

If this is true, then the software patch could be used to create ghost entries into Aadhaar database. The new IDs created in UIDAI may also be misused to siphon off rations meant for poor.

How did the hack happen?

In 2010, the UIDAI allowed private agencies to enrol users to the Aadhaar system in order to speed up enrolments. In the same year, Bengaluru-based Mindtree won a contract to develop an official, standardised enrolment software – called the Enrolment Client Multi-Platforma  (ECMP )- that would be installed onto the thousands of computers maintained by these private operators.

Instead of using a web-based system in which all software would be installed on the UIDAI’s own servers and enrolment operators would have a user name and password to access the system, softwares were installed on each enrolment computer.

According to the report, B. Regunath, a software architect who led the team at Mindtree that worked on the project, said a web-based enrolment software for Aadhaar was not practical at the time because many parts of the country had very poor Internet connectivity.

To make data security foolproof, more features were added to the software that was used by Aadhaar enrolment operators. They were required to log in to the software by first providing their own fingerprint or iris scan. Also, a GPS device was attached to verify the location.

However, the report claimed that in early 2017, these security features were bypassed by a software hack. There is also a video on YouTube which offers a step-by-step guide to bypass these security features.

UIDAI’s response

Unique Identification Authority of India (UIDAI) has dismissed the claims as completely baseless. Refuting the allegations, UIDAI said no operator can make or update Aadhaar unless an individual gives biometrics details.

The UIDAI said that it matches all the biometric – 10 fingerprints and both iris – of a resident enrolling for Aadhaar with the biometrics of all Aadhaar holders before issuing an Aadhaar, and so the claims of introducing information into Aadhaar database were “completely unfounded”.

The post Big data breach! Aadhaar software hack raises major security concerns appeared first on Artificial Intelligence.