A disturbing trend of sophisticated online radicalisation has emerged, with a royal commission hearing that far-right individuals are deliberately deploying memes and coded language to propagate “more palatable” antisemitism, sidestepping digital platform safeguards.
This insidious tactic allows extremist narratives to permeate mainstream online spaces, making their detection and removal increasingly difficult for social media moderators. The revelations underscore a growing challenge for authorities seeking to curb the spread of hate speech in the digital age.
The Cloaked Language of Hate
Far-right groups are not just sharing explicit hateful content; they are increasingly adopting internet memes – often humorous or relatable images, videos, or pieces of text – to embed antisemitic messages within a seemingly harmless context. This strategy leverages the viral nature of memes and their capacity for ambiguity, making it harder for automated moderation systems, and even human moderators, to discern malicious intent. ABC News Arts & Lifestyle reported on the findings presented to the commission, highlighting how this 'memeification' of hate allows extremist viewpoints to spread under the radar.
Experts presenting to the royal commission detailed how these coded messages often draw on established antisemitic tropes but present them in new, disguised forms. For instance, instead of direct threats, a meme might subtly allude to conspiracies involving Jewish individuals or groups, wrapped in an ironic or humorous package. This makes the content appear less overtly hateful, thus reducing the likelihood of it being flagged for violating community guidelines. The goal is to normalise these harmful ideas, drip-feeding them into online conversations until they become part of accepted discourse, particularly among younger, digitally native audiences.
The Moderation Maze: A Game of Cat and Mouse
The use of coded language and memes presents a significant hurdle for social media companies. Their automated systems, often reliant on keyword detection and known hate symbols, struggle to interpret nuanced or ironic expressions of extremism. When human moderators intervene, they face an immense volume of content and the challenge of understanding the often-obscure cultural references embedded in these memes. This creates a perpetual game of cat and mouse, where extremists continuously adapt their tactics to outmanoeuvre platform safeguards.
The cost of this sophisticated evasion strategy extends beyond the immediate challenge for content moderators. It represents a substantial financial and reputational burden for platforms, which are under increasing pressure from governments and the public to combat online hate effectively. Furthermore, the spread of such content, even in its 'palatable' form, contributes to the normalisation of antisemitism and poses a genuine threat to social cohesion within Australian communities.
Impact on Australian Communities
The findings presented to the royal commission are particularly pertinent in Australia, where concerns about rising antisemitism and far-right radicalisation have been prominent. The ability of these groups to operate with a degree of impunity online, by camouflaging their hateful messages, means that vulnerable individuals and communities are exposed to insidious forms of discrimination. While the full extent of the impact on Australian communities is still being assessed, the implications for social harmony and the safety of minority groups are undeniable.
Community organisations and anti-hate groups have long warned about the evolving nature of online extremism. They argue that platforms must invest more in culturally sensitive and linguistically adept moderation teams, alongside developing more sophisticated AI tools that can detect contextual nuances rather than just explicit keywords. The challenge lies in balancing freedom of expression with the imperative to protect individuals from hate speech and incitement to violence – a Tightrope walk that requires constant innovation and vigilance from all stakeholders.
