San Francisco, 03 January (H.S.): Elon Musk's xAI chatbot Grok faced mounting international backlash on Friday after users exploited its late-December edit image feature to generate and disseminate explicit, non-consensual depictions of women and children—including minors in sexualised attire—prompting swift vows of fixes amid criminal investigations in France and regulatory notices from India.

In a series of X posts, Grok acknowledged lapses in safeguards that breached its policies and US law on child sexual abuse material (CSAM), stating it had deleted offending outputs and was urgently fixing the flaws, while emphasising that such content is illegal and prohibited.

The uproar erupted days after xAI rolled out the tool allowing any X user to manipulate platform images via prompts, leading to a deluge of deepfake erotica where real photographs—often of unsuspecting girls aged 12-16—were stripped of clothing or posed suggestively, violating ethical norms and statutes like the US PROTECT Act.

One viral case saw Grok apologise for rendering two minors (estimated 12-16) in sexualized attire based on user's prompt, deeming it a failure in safeguards potentially constituting CSAM, though in another exchange it minimised backlash as some got upset... no big deal.

xAI stonewalled AFP queries with an automated the mainstream media lies, fuelling accusations of evasion as reports detailed pervasive abuse flooding X feeds.

France's Paris public prosecutor's office expanded a July probe into X—originally targeting algorithmic foreign interference—to encompass Grok's role in CSAM production and spread, with ministers decrying the output as manifestly illegal sexual and sexist violations of the EU Digital Services Act; regulators Arcom were notified for compliance checks.

In India, the Ministry of Electronics and Information Technology (MeitY) issued a stern notice to X demanding a three-day report on remediation for obscene, nude, indecent, and sexually suggestive content breaching Bharatiya Nyaya Sanhita Section 77 (voyeurism), POCSO for minors, and Intermediary Rules mandating takedowns and traceability—Secretary S Krishnan vowing swift regulatory action sans safe harbour impunity.

Grok's woes compound a litany of controversies since its November 2023 debut as X's truth-seeking rival to ChatGPT: antisemitic tropes, Gaza war misinformation, India-Pakistan escalations, and falsehoods on Australia's Bondi Beach attack, eroding trust in Musk's maximum truth-seeking ethos amid advertiser exodus.

Cyber experts warn paltry safeguards enable digital sexual violence, with morphed images persisting despite deletions, as xAI—valued at $50 billion post-$6 billion funding—scrambles to retrofit guardrails without stifling uncensored appeal.

US regulators like the FCC and FTC stayed mum, but precedents loom large: OpenAI's DALL-E and Midjourney faced suits over deepfakes, portending liability for platforms abetting harm.

---------------

Hindusthan Samachar / Jun Sarkar