Are Palash Muchhal’s viral chat screenshots real? We asked ChatGPT to verify cheating speculations

Online rumours claim Palash Muchhal cheated on Smriti Mandhana even though nothing has been confirmed by either of them. Screenshots shared on Reddit and Instagram allegedly show him flirting with a woman named Mary D’Costa

He allegedly invited her for a swim, a spa outing and an early morning walk on Versova beach. The chats also mention him allegedly saying that his long-distance relationship ‘s*cks’ and that they meet only once in three to five months.

We asked ChatGPT to verify the authenticity of the viral chat screenshots, which claim that Palash Muchhal allegedly cheated on Smriti Mandhana. It apparently involves a woman named Mary D’Costa.

The artificial intelligence (AI) agreed to analyse it “purely from a technical and behavioural point of view”. “This is not a verdict since neither Smriti Mandhana nor Palaash Muchhal has confirmed anything,” it said.

Why it could be real

According to ChatGPT AI, certain signs make the screenshots appear more realistic. First, the Instagram DM layout, fonts, timestamps, reactions and icons match the current design. There are no obvious visual glitches that usually appear in edited chats.

Secondly, the flow of conversation feels natural, with casual language, typos, pauses, emojis and quick emotional tone changes, which are hard to script perfectly.

Third, the screenshots show continuity across multiple pages, repeatedly mentioning the same topics like swimming, Versova and Silver Beach, spa plans and “long-distance relationship being dead most of the time”.

Finally, message stacking and emoji reactions look technically accurate and not copy-pasted.

Why it could be fake

At the same time, ChatGPT flags some signs that suggest the leaked screenshots might not be genuine. The Instagram account that shared them, marydcosta_, is not a verified person or a public figure. So, her credibility cannot be checked, the AI chatbot says.

Only still images were posted, and not a screen recording. A video would have been stronger proof because photos are easier to edit, according to ChatGPT.

The chats show only one person’s phone, so anyone could theoretically create a mocked conversation on their own screen.

ChatGPT also suspects the timing of the viral post as the screenshots surfaced only after the wedding had been postponed, which “conveniently amplifies public drama”. Such patterns often occur in manufactured controversies, it adds.

“Even if screenshots are real, private chats leaked without consent are still a privacy breach, not legal proof of cheating and not justification for harassment toward anyone involved,” says ChatGPT.

During 24-25 November, the interest in ‘Smriti Mandhana’ soared on Google India: