Posted: 18 December 2025 • Category: Society, Technology, Public Policy
Bondi, Hate Speech Crackdowns, and the Problem We Keep Missing
If we only treat the words, we miss the furnace that forges them.
In the days since the Bondi tragedy, Australians have demanded action. Government has signalled tougher laws aimed at “hate preachers”, serious vilification, and stronger penalties where hate promotes violence. The intention is understandable: draw a line, send a message, and protect communities.
I agree with the message. But I’m not convinced we’ve correctly identified the main problem. Hate speech is often a symptom — not the cause. It’s a signal that someone’s worldview is becoming dangerous, and that the social environment around them (online and offline) is feeding that deterioration.
Thesis: A tough stance may be necessary in the interim. But the long-term solution is upstream: communication standards, evidence-based norms, education, and embedded mental health support.
The timeline matters: the attack, then the online wave
The attack occurred on Sunday 14 December 2025. Many Australians experienced the shock in real-time or woke into the aftermath on Monday 15 December — the day tributes began to gather and the national conversation ignited.
And almost immediately, a second wave arrived: the online wave. Not simply news updates — but social media “explanations”, blame narratives, recruitment energy, and emotion-driven certainty. This is the part we keep underestimating: mass violence is not only an event. It becomes an idea-object online — something people weaponise for identity, tribal belonging, and ideological theatre.
Hate speech is hard to police — and that’s not the whole story
Regulators and safety bodies rightly treat online hate as serious. In the most extreme cases, it correlates with real-world harassment and violence. But the practical reality is brutal: content volumes are enormous, context is slippery, and virality outruns moderation.
So yes — law enforcement and platform enforcement matter. But if our main lever is “remove the worst posts”, we’re trapped in an endless mop-up operation, always arriving after the harm has already spread.
Here’s the risk: crack down on speech, and you drive instability underground
This is the part nobody wants to say out loud: if you clamp down purely at the “speech” layer, you don’t remove the underlying belief system. You often force it into the dark.
That means private groups, fringe platforms, and encrypted channels. And once the public conversation moves there, the next political pressure wave tends to become: “we must break encryption to stop extremism.”
Australia already has “lawful access” frameworks and ongoing debates about encryption and compelled assistance. If we let this slide into a broad “encryption war”, we risk weakening privacy and security for everyone — while still failing to treat the upstream conditions that generate hatred.
Hate speech can be a smoke alarm: it shows who needs help
I’m not romanticising hate speech. I’m saying something more uncomfortable: when someone starts pouring out identity-based hostility, it can be a visible indicator of deterioration. A person is being shaped by grievance loops, isolation, paranoia, obsession, or radical ideology — sometimes all at once.
In those cases, hate speech is not merely “content to be deleted”. It’s a warning signal: this person is becoming unsafe — for others, and often for themselves.
If we erase every signal without building an off-ramp, we may simply end up with a society that is more blind, not more safe.
What we should do instead: hold the line, then go upstream
1) A tough stance in the interim
There must be clear consequences for incitement, intimidation, and organised promotion of violence. A firm response tells communities: you are protected, and it tells would-be agitators: this is not tolerated. That message matters.
2) Then shift focus to the real battleground: the fermentation stage
The central problem isn’t only what people say. It’s how online systems ferment belief long before speech becomes prosecutable. We need upstream infrastructure:
-
Embedded mental health support: not only hospitals and crisis lines, but local community access points, outreach teams, culturally competent services, and fast pathways for families who notice fixation and escalation.
-
Education for evidence-based thinking: media literacy, emotional regulation, conflict skills, basic logic and fallacy detection, and “how algorithms shape you” training.
-
Off-ramps from radicalisation: treat grievance loops like a public-health issue as well as a security issue. Some people need intervention, not only punishment.
The missing piece: communication standards for the digital public square
We already know how to run shared communication spaces. Consider amateur radio: it operates under rules, identification norms, and procedures — not to “ban speech”, but to keep a public channel usable and safe.
Social media is a public channel too — except it’s supercharged by recommendation algorithms and viral distribution. It functions as a town square, a broadcast station, and a personal billboard all at once. It is also a marketplace for identity and ideology — even when no money changes hands.
Simple idea: If we can apply standards to advertising because ads shape behaviour, we can apply standards to algorithmic public communication because it shapes behaviour too. Ideas carry “currency”. And sometimes that currency buys real-world harm.
A practical standards model (without turning into thought police)
-
Accountability without doxxing: reduce disposable anonymity at scale while preserving privacy for ordinary users.
-
Claim hygiene: prompt users (especially high-reach accounts) to distinguish observation, inference, and opinion. Don’t ban emotion — stop rewarding unsupported certainty.
-
Circuit breakers during crises: slow resharing, require “read before repost”, limit forwarding, and reduce algorithmic amplification of unsourced inflammatory content when tensions are high.
-
Transparency obligations: publish moderation metrics, response times, and systemic risk assessments. If platforms are public infrastructure, they should be auditable like public infrastructure.
-
Status for evidence-based speech: make it culturally cool to show your working. Evidence should be the high-status behaviour, not pontification.
The bottom line
Strong laws can draw boundaries. But boundaries don’t cure what grows inside them. If we want fewer tragedies, we must treat hate speech like a smoke alarm: it’s a warning signal — and a call to intervene earlier, not merely delete louder.
Hold the line now. But immediately begin the real work: mental health embedded in communities, education that trains evidence-based thinking, and communication standards that stop platforms monetising emotional sabotage.
References & further reading
