Evaluating Anthropic’s Credibility in Cybersecurity: A Community Perspective on Trust and Transparency

The Cybersecurity Community’s Trust in Anthropic: A Shifting Landscape

The cybersecurity industry relies heavily on trust. Vendors must demonstrate their commitment to responsible practices, and customers must believe in their ability to adhere to these principles. Anthropic, a prominent player in the AI landscape, has been a shining example of a company prioritizing trustworthiness. However, recent events have raised questions about the company’s dedication to its values.

Shifting Commitment to Trustworthiness

In January 2026, Anthropic launched a series of ads criticizing OpenAI’s approach to user data, touting its own commitment to user privacy. This move was seen as a bold statement of its values. However, just a few weeks later, Anthropic released an updated version of its Responsible Scaling Policy (RSP), which marked a significant shift in its approach to safety and risk management.

Relative Approach to Safety and Risk Management

The original RSP committed to maintaining absolute risk levels below acceptable thresholds, regardless of competitor behavior. In contrast, RSP 3.0 adopts a relative approach, where Anthropic’s safety commitments are now contingent on the actions of its competitors. This change has raised concerns among security professionals, who worry that the company’s commitment to safety is no longer absolute.

Concerns and Uncertainty

Meanwhile, rumors of a rift between Anthropic and the Pentagon surfaced, culminating in the company’s termination as a contractor. Anthropic’s CEO, Dario Amodei, had previously stated that the company would not work with the Pentagon unless it agreed to certain conditions, including not using its models for mass surveillance or autonomous weapons. The Pentagon’s decision to replace Anthropic with OpenAI, which later backtracked on its initial commitment to uphold Anthropic’s red lines, has further complicated the situation.

Evaluating Trustworthiness

The events of the past six weeks have raised questions about Anthropic’s trustworthiness. While the company’s initial commitment to responsible practices was laudable, its recent actions have introduced ambiguity into its relationships with customers and partners. The cybersecurity community is now faced with the challenge of evaluating whether Anthropic’s actions align with its values.

Implications for the Cybersecurity Community

As AI models become increasingly integrated into code review, vulnerability management, and security architecture, the stakes are higher than ever. The industry needs vendors that can demonstrate unwavering commitment to responsible practices. Anthropic’s RSP 3.0 has introduced a level of relativity that undermines its previous commitment to absolute safety standards.

About Author

Vaibhav

See author's posts