High Court
Delhi HC quashes tax notices against Prannoy Roy & Radhika Roy, fines department Rs 2 Lakh
NEW DELHI: In a sharp rap on the knuckles for tax overreach, the Delhi High Court has told the Income Tax Department that it cannot keep knocking on the same door hoping for a different answer, especially when it has already been opened, inspected and firmly shut.
Quashing reassessment notices issued to veteran broadcaster Prannoy Roy and media professional Radhika Roy, the court on January 19 ruled that the tax authorities had acted without jurisdiction, reopening a settled assessment on nothing more than a change of opinion. To underline its displeasure, the court imposed a token cost of Rs 1 lakh each, Rs 2 lakh in total, on the department, payable to the Roys.
The case, like a badly written sequel, centred on Assessment Year 2009–10, an old chapter the tax department tried to reread years later.
Radhika Roy had filed her income tax return for AY 2009–10 on July 31, 2009, declaring an income of Rs 1.66 crore. The return was processed and accepted under Section 143(1), with the intimation issued on February 22, 2011.
Then came the first knock. In July 2011, the department reopened the assessment under Sections 147 and 148, citing transactions involving shares of New Delhi Television Ltd (NDTV) between the Roys and their holding company, RRPR Holding Pvt Ltd. The reassessment culminated in an order dated March 30, 2013, assessing Radhika Roy’s income at Rs 3.17 crore. This included a major addition of Rs 1.30 crore as short-term capital gains, along with smaller additions of Rs 20.74 lakh as house property income and Rs 2,750 relating to Section 80G.
Crucially, during these proceedings, the assessing officer had specifically examined interest-free loans received by the Roys from RRPR. A show-cause notice issued on March 6, 2013 proposed treating these loans as “deemed dividends” under Section 2(22)(e). After examining RRPR’s audited books, balance sheets and shareholding pattern, the officer dropped the proposal. No addition was made on this count.
Three years later, on March 31, 2016, the department reopened the same assessment yet again, issuing fresh notices under Section 148 to both Prannoy Roy and Radhika Roy. This time, the department leaned on “complaints” and an internal review of RRPR’s records, arguing that interest-free loans given to the Roys should be taxed as “deemed income” under Section 2(24)(iv).
The figures were hefty. RRPR had borrowed Rs 375 crore from ICICI Bank in October 2008 at an interest rate of 19 per cent per annum. From this loan, it extended interest-free advances of Rs 20.92 crore to Prannoy Roy and Rs 71 crore to Radhika Roy. According to the department, RRPR suffered interest costs of nearly Rs 35 crore in that year, and an estimated Rs 6.79 crore of “benefit” had accrued to Radhika Roy alone due to non-charging of interest.
A bench of justices Dinesh Mehta and Vinod Kumar held that the so-called “new information” was neither new nor hidden. The interest-free loans were already disclosed, examined and consciously accepted during the earlier reassessment proceedings.
“Section 147/148 powers are an exception, not a licence for repeated harassment,” the court observed, noting that the same transaction cannot be reopened merely because a different officer believes another legal provision should have been applied.
Calling Sections 2(22)(e) and 2(24)(iv) “two sides of the same coin”, the court said the department had every opportunity in 2013 to tax the alleged benefit if it believed it was taxable. Revisiting the issue years later was nothing but a change of opinion, a settled no-go zone in tax law.
The court also rejected the department’s attempt to invoke the extended six-year limitation period by alleging failure to disclose material facts. The Roys, it said, had disclosed all primary facts, including RRPR’s audited accounts, which explicitly recorded the interest-free loans. Drawing on Supreme Court precedents, the bench reiterated that an assessee is not required to disclose inferences or help the tax officer draw conclusions.
Allowing both writ petitions, the High Court quashed the 2016 notices and all consequential proceedings. While noting that “no amount of cost can be treated enough” for such cases, it imposed Rs 1 lakh as cost in each petition, a symbolic but pointed message.
Beyond the Roys, the ruling sends a wider signal. Reassessment powers are not a rewind button. Once the taxman has examined the facts, applied his mind and passed an order, he cannot keep returning with fresh labels for the same transaction.
In short, the court told the department to stop re-editing old tapes, especially when the credits have already rolled.
High Court
Bombay High Court questions AI celebrity deepfakes in Shilpa Shetty case
Justice questions legality of unconsented AI personas, platforms directed to respond.
MUMBAI: The Bombay High Court just put AI on the witness stand because when a chatbot starts chatting as Shilpa Shetty without asking, even the bench wants to know who gave permission. The Bombay High Court on Wednesday expressed serious concerns over the legality of artificial intelligence tools that simulate celebrity personalities without consent, during a personality rights suit filed by actor Shilpa Shetty.
Justice Sharmila Deshmukh, hearing the matter, questioned platforms that allow users to interact with AI-generated versions of actors without authorisation. The court noted that one accused AI chatbot website continued using Shetty’s personality without permission, prompting the judge to ask about the legal basis for such operations.
When the lawyer for the AI company argued that the system relied on algorithms and did not require celebrity consent, Justice Deshmukh challenged the platform’s right to recreate and make public a person’s identity in this manner. She observed that while users uploading photographs raised one set of issues, AI systems generating content based on recognised personalities posed distinct legal and ethical questions especially when the platform itself acknowledged the content was not real.
The court directed the platform to file a detailed response explaining its position.
The case involves Shetty seeking restrictions on more than 30 platforms including e-commerce websites and AI services accused of hosting or enabling misuse of her image and circulation of deepfake content.
The Bench also raised concerns about Youtube commentary videos discussing the ongoing proceedings involving Shetty and her husband, questioning whether unverified discussions could malign parties without journalistic checks.
Counsel for Google, Tenor and the AI entity informed the court that flagged infringing URLs had been removed. Shetty’s team disputed this, leading the court to allow her to file an application alleging non-compliance if links remained active.
Tenor objected to the broad injunction sought, arguing it functions as an intermediary GIF platform without capacity for proactive monitoring. The court directed Tenor to file an affidavit opposing the order.
E-commerce platforms including Amazon stated they had removed unauthorised listings using Shetty’s name and image, and would continue to act on specific notifications.
The court reiterated that directions for intermediaries would operate on a “take-down on notice” basis, requiring removal of infringing content once flagged.
As deepfakes blur the line between real and rendered, the Bombay High Court isn’t just hearing a case, it’s asking the bigger question: in the age of AI avatars, who really owns your face?
