In 2020, key members of OpenAI left to form Anthropic, a public benefit corporation with a laser focus on AI safety. Anthropic believes that AI has the potential to radically change society, and that ethical standards are required to minimize the dangers and maximize the benefits. They act on this mission in two ways. First, by conducting research that reveals the oppor­tu­ni­ties and risks of AI, which will inform future policy for its safe use. And second, by building reliable, inter­pretable, and steerable AI models that set the gold standard for safety. In mid-2023, that research culminated in the release of Claude, a ground­break­ing virtual assistant, and Anthropic’s first product release.

Over the course of two and a half years we worked with Anthropic, from stealth mode to product launch, helping them build a brand infused with their values, that stands apart in a category rife with tech tropes.

Partnering early on with company founders, we designed a visual identity to bring them out of stealth with credibility and a clear sense of what makes them different. As encap­su­lat­ed by their name, everything Anthropic does is focused on aligning technology to human values. It was important that the identity captured at once the company’s human-centered mission, and a sense of the technical craft with which they do their work.

We began with the foun­da­tion­al visual elements. We opted for a pure, typographic logo with a single standout detail. The slash is a reference to the code that underlays AI, and a nod to the future ahead. We introduced a typographic palette — a combination of the Styrene family by Commercial Type, and Klim’s Tiempos family — that is both technically refined and charmingly quirky. And we developed a color system that would bring warmth to the brand while being able to handle the dual needs of marketing com­mu­ni­ca­tions and product UI.

Later, we completely overhauled Anthropic’s digital presence in preparation for their big coming-out moment and the launch of their first publicly available product: Claude.

Chat GPT, LLaMa, Bard… If the first wave of Big Tech AI products were built for human consumption, they sure don’t sound like it. We wanted to frame Claude in a unique way, grounded in its difference: Claude was built with Con­sti­tu­tion­al AI, a training method that makes it safer, more reliable, and less likely to give wrong answers than its con­tem­po­raries. It was designed to be helpful, harmless, and honest always.

Our interactive work focused on translating these principles into design, helping Anthropic talk about Claude and the company behind it while expanding their foun­da­tion­al visual identity into a more robust digital brand toolkit. We col­lab­o­rat­ed with company stake­hold­ers to develop the site’s archi­tec­ture, design, and messaging. We created an illus­tra­tion language that doubles down on the human values that drive the company, and allows them to talk about their work in a way that is consumable for both technical and general audiences alike. We brought in our friend and SF-based pho­tog­ra­ph­er, Aaron Wojack, to visually tell the story of the company and its culture. We developed an extensive UI component system that puts function first without losing the soul of the brand, and that works equally well in web and product appli­ca­tions. And we brought all of this to life with a modular CMS, engineered for maximum flexibility as the company continues to scale and release new products.

“Geist were instrumental in helping us define our brand identity.”

— Jack Clark, Co-Founder

We’re proud to have played a small part in the initial success of Anthropic. Since the launch of the site Anthropic has raised more than $1.4 billion in funding, and has formed strategic part­ner­ships with brands like Google, Salesforce, and Zoom.

Nobody yet knows just how big an impact AI will have on the ways we live and work. While we believe in the positive potential of this technology, we’re equally sober about its risks. Our ambition is to use design to make novel tech­nolo­gies more useful, under­stand­able, and meaningful. We’re happy to align ourselves with experts in this emerging space who are motivated by a vision for the common good.