Every day, millions of people scroll through social media, often unaware that hate hides behind images, memes, or seemingly harmless symbols. We live in a digital era, profoundly shaped by social media. And it is precisely in these spaces that we witness a significant transformation in the language of hate. Increasingly, hate is no longer expressed openly or directly. It no longer appears simply as denial or blatant attacks. Instead, it manifests through trivialization, banalization, and normalization of historical atrocities or the marginalization of vulnerable communities.

What is Coded Language?

A key mechanism in this evolution is coded language: a way of expressing hate that conceals harmful intent behind symbols, numbers, emojis, or seemingly innocuous images. In antisemitic discourse, for example, numbers might replace letters, emojis or symbols may allude to gas chambers, and seemingly harmless and ordinary images are used to reference Jews, the Holocaust, or its victims.

These examples are cited not to normalize them, but to make them visible and highlight the urgent need to confront them. Coded language is deliberate, strategic, and systematic. It allows hate to spread quickly, often escaping immediate moderation, gaining visibility through likes, shares, and comments.

The Harm Behind the Codes

Coded language is far from harmless. In using coded language, historical crimes are stripped of their meaning, trauma is distorted into entertainment, and memory and victims are turned into objects of ridicule. Often, this happens through trivialization: turning atrocities or discrimination into jokes, memes, or casual references. This makes hate appear harmless or “normal,” lowering the social and moral barriers to engagement. Even when such content slips past moderation, it inflicts real harm: it normalizes dehumanization, erases historical responsibility, and fosters a culture of hate.

A new dimension of concern is also the use of artificial intelligence to create false or manipulated images. AI-generated content can depict historical events inaccurately or fabricate scenes that never happened, often targeting vulnerable communities. In the case of the Holocaust, for instance, AI-generated images can distort visual memory, trivialize atrocities, or even promote denial. The impact is profound: when false visual narratives circulate online, they undermine collective memory, erode trust, and amplify the harm caused by coded language.

Technology, Algorithms, and Institutional Challenges

The spread of coded hate is amplified by the very technologies meant to connect us in various ways:

• Viral diffusion: social media algorithms reward engagement, which unintentionally boosts posts containing subtle or coded hate, often packaged as humor or memes, a form of trivialization that lowers users’ critical awareness and increases the likelihood of engagement.

• AI and deepfake content: manipulated images and videos can distort historical events or target marginalized groups, making misinformation highly persuasive.

• Automated moderation limits: current moderation tools struggle to detect subtle codes, symbols, evolving slang, or the use of trivialization as camouflage, creating gaps in protection.

For institutions, these dynamics present serious challenges. Legislators and regulators must navigate the tension between free expression and accountability, while platforms are pressured to develop more sophisticated moderation systems. At the same time, there is growing recognition of the potential for technology to be reoriented: AI and algorithms could be harnessed to detect, flag, and counter coded hate, rather than inadvertently amplifying it.

Confronting Coded Hate

Recognizing coded language and its digital evolution is critical. It is not harmless or abstract; it is a deliberate strategy to bypass accountability while perpetuating harm. In both antisemitic and anti-LGBTQ+ contexts, coded language works to erode empathy, distort reality, and weaken the social consensus that protects marginalized communities.

Awareness is the first step. By identifying these codes, understanding their meanings, and speaking out, we can begin to disrupt the silent spread of hate online. Digital spaces should not be arenas where prejudice hides behind a mask of humor or anonymity. They must be places where memory is respected, diversity is protected, and accountability is upheld.

It is everyone’s responsibility to recognize and report coded hate. Without reporting, these incidents effectively do not exist, and policymakers cannot take action or implement measures. Only by acting together can digital spaces become safe, inclusive, and respectful of all communities.