New Delhi, February 27 (IANS). The increasing use of Artificial Intelligence (AI) during education in schools has given rise to a new debate. This discussion is no longer limited to just making education easy and effective, but has also reached the safety of children and their rights. Especially the threat of possible misuse of students’ data is emerging as a serious and worrying reality at the global level.
A group of UN departments, including UNRC, UNESCO, UNICRI and the Office of the UN High Commissioner for Human Rights, issued a joint statement. The statement stressed the need for security and privacy of children’s data collected from AI systems, tools and platforms.
In the year 2025, a lawsuit was filed against PowerSchool, a leading edtech company based in Texas, USA, regarding a major data breach. The company is accused of leaking sensitive information of more than 6 crore students and 1 crore teachers in this incident, which in many cases included personal details like social security numbers. PowerSchool works as a student information system.
A pilot study conducted in India has also presented worrying pictures. According to this, Indian educational institutions faced more than two lakh cyber attacks and nearly four lakh data breach incidents in just nine months.
This global and national perspective is important to understand the recently announced partnership between Pratham and Anthropic. Pratham is a non-governmental organization active in the field of education, while Anthropic is a US-headquartered AI company that develops artificial intelligence systems.
Under this partnership between the two organizations, Anthropic’s large language model ‘Cloud’ gets access to children’s handwritten responses, data related to their academic progress and the information needed to create personalized feedback.
Pratham-Anthropic Partnership together announced their first product ‘Anytime Testing Machine (ATM)’ in February 2026. Powered by Anthropic’s cloud, ATM creates curriculum-aligned tests. Digitizes student’s handwritten answers. Additionally, it applies rubric-based grading using Cloud and provides bilingual (Hindi-English), personalized feedback. Given India’s DPDP Act, there may be some risks in the assessment model.
Section 9(1) of the DPDP Act mandates that a data trustee must obtain verifiable consent of a parent or legal guardian before processing any personal data of a child (defined as anyone under the age of 18 years).
The Draft DPDP Rules, 2025 (Rule 10) provide further guidance on the mechanism for obtaining such consent, including OTP-based parental consent and integration with government-issued IDs.
However, in the case of children’s data being used by ATMs, parents may not fully understand that their child’s handwritten work is being photographed, uploaded to a cloud-based AI system, processed by a US company’s servers and analyzed by a large language model.
–IANS
KK/ABM
