Anthropic and AI Safety Marketing: A Trillion-Dollar Empty Slogan?

Anthropic and AI safety marketing: trillion-dollar empty slogan?

In the Habr section dedicated to artificial intelligence, a column was published with harsh criticism of Anthropic's marketing rhetoric around "AI safety." The author of the material argues that Anthropic's public image as a company taking a responsible approach to AI development differs in practice almost nothing from the behavior of other major market players — and all the rhetoric about safety boils down to an empty slogan.

What is the essence of the author's criticism

The central thesis of the column is formulated directly: if you look not at the words of companies, but at their real actions, there are very few differences between Anthropic and other major AI developers. The author assigns Anthropic to the same group as the rest of the major tech players, calling all of them one term — "the caste of techno-oligarchs," referring at the same time to external material with a more detailed analysis of this concept. In the author's view, the rhetoric about AI safety and ethics is first and foremost an element of market positioning, not a reflection of a fundamentally different business model or values.

Anthropic initially positioned itself as a company founded by OpenAI veterans precisely because of disagreements over the pace and safety of AI development — and for years has maintained the image of a more cautious and responsible player compared to competitors. It is this contrast between the company's founding story and its current behavior in a highly competitive market that most often becomes the focal point for such criticism: the stronger the declared mission differs from typical technology business behavior, the more noticeable becomes any discrepancy between words and deeds.

• Publication platform — Habr, section dedicated to AI.
• Format — author's column with personal assessment, not a news item.
• Main thesis — the author calls Anthropic's "AI safety" marketing an empty slogan.
• Comparison — according to the author's formulation, Anthropic and other major AI companies belong to the same "caste of techno-oligarchs."
• Explanation of the utility of AI products — the author links it not to company ethics, but to the absence of a monopoly in the market.

What role does the author assign to users

Separately, the column formulates a harsh thesis about the place of the ordinary user in this system: in the author's opinion, companies need people in exactly two roles — as suppliers of behavioral data used for training and improving models, and as consumers of finished products based on these same models. From this framework, the author derives an explanation of why AI services ever turn out to be useful, despite regular errors and "hallucinations": the only reason, in the columnist's opinion, is that no single company has a monopoly in the AI market, and therefore competition forces all players to make a product of sufficient quality in order not to lose users to alternatives.

Why such columns find resonance with readers

Such criticism fits into a broader and long-standing discussion around so-called "safety-washing" in the AI industry — accusations that public rhetoric about risks and ethics is used by companies as a tool for positioning and regulatory lobbying, rather than as a real constraint on their own commercial decisions. Anthropic in this discussion becomes a frequent target precisely because AI safety has historically occupied a central place in the company's public positioning.

At the same time, the column itself does not offer specific verifiable evidence of violations by Anthropic — the material is built as a journalistic statement, not as an investigation with specific facts. This does not negate the fact that such texts reflect real sentiments of a part of the technical community, tired of the contrast between companies' declarations of responsible AI and the usual race for market share and capitalization in the industry.

Such texts also often become a platform for broader reflection on the role of large technology business in the life of an ordinary user: if the value of a product for a person is determined not by the company's concern for his wellbeing, but exclusively by competitive pressure in the market, then the very category of "ethical AI" in market rhetoric turns out to be more of a tool for brand differentiation than an independent, independently verifiable product characteristic.