OpenAI Alleges Israeli Firm Tried to Disrupt Lok Sabha Polls with Anti-BJP Agenda

An OpenAI report has revealed an Israel-based network that generated content critical of India's BJP and supportive of Congress. The activity, flagged in May, focused on Indian elections. OpenAI claims it disrupted covert operations attempting to use AI models to sway elections, just days before Lok Sabha poll results are announced.

2>Covert Operations Unveiled

The OpenAI report identifies a for-hire Israeli firm, STOIC, generating comments criticising BJP and praising Congress. This activity targeted Indian polls and was flagged in May. The network used AI to manipulate public opinion and influence political outcomes, attempting to deceive people about their identities and objectives.

The report claims a cluster of accounts, operated from Israel, were used for these covert operations. Content was shared on X, Facebook, Instagram, websites, and YouTube. "In early May, it began targeting audiences in India with English-language content," the report states.

OpenAI, founded in December 2015, is an artificial intelligence research organisation. Reacting to the report, Union Minister of Electronics and Technology Rajeev Chandrasekhar commented on the influence operations targeting BJP. He called it a "dangerous threat" to democracy.

Official Reactions

Rajeev Chandrasekhar stated, "Vested interests in India and outside are driving this and it needs to be scrutinised, investigated and exposed." He criticised the timing of the report's release, suggesting it should have come out earlier, not as elections ended.

China-Based Network Exposed

Meta also released a report about a China-based network promoting a pro-Khalistan agenda. The network targeted the Sikh community in India, Australia, Canada, the UK, New Zealand, Pakistan, and Nigeria. The operation used compromised and fake accounts to pose as Sikhs, post content, and manage Pages and Groups.

The report claims a fictitious activist movement, called Operation K, was created by the network, calling for pro-Sikh protests. Meta removed this activity early, preventing it from building an audience among authentic communities. The network posted in English and Hindi about news and used images manipulated by photo editing tools or generated by AI.

Content flagged by Meta included topics on floods in Punjab, the Sikh community worldwide, the pro-Khalistan movement, the killing of terrorist Hardeep Singh Nijjar in Canada, and criticism of the Indian government. The Meta report highlights the use of compromised accounts to push these narratives.

Published On: May 31, 2024