Grok, guardrails, and the cost of ‘move fast’ AI

Today, ICMEC Australia CEO, Colm Gannon has vocalised concerns following the joint press conference by the Prime Minister, Minister for Communications, and the eSafety Commissioner - warning that Australia is entering a defining moment for the safety of children and women in technology. 

His new op-ed (attached here), examines the global fallout from the deployment of Grok, X’s generative AI chatbot, after regulators began receiving reports of non-consensual sexualised imagery involving children and women and other exploitative outputs. 

Australia’s eSafety Commissioner this week commenced a second investigation into X, coinciding with new mandatory online safety codes coming into force in March 2026 — a shift Mr Gannon says will reshape expectations for how AI systems are designed, governed, and deployed. 

At the same time, regulators across Europe, the UK, and parts of Asia have issued bans, investigations, and legal orders relating to Grok, signalling what Gannon describes as a decisive end to the industry’s “deploy first, fix later” culture. 

Key takeaways from Colm’s op-ed: 

• The Grok controversy reflects a platform governance and safeguarding failure — not a technical glitch. 

• Generative AI must anticipate misuse, rather than repair harm after it occurs. 

• Neutrality is no longer defensible when misuse is foreseeable in product design. 

• User blame is obsolete in cases where harm was predictable from the outset. 

• Australia’s forthcoming safety codes will set the benchmark for future AI deployment and accountability. 

Colm Gannon's Op-ed: Grok, guardrails, and the cost of ‘move fast’ AI

-Ends-