Picking an AI brand ambassador

The AI you're building might sabotage your brand strategy.

A new study from Oxford and King's College tested AI models from Google, OpenAI, and Anthropic in 32,000 strategic decisions. What they found might change how you think about choosing your company's AI partner.

The Setup
Each AI faced classic prisoner's dilemmas with varying time horizons. Cooperate and both parties win a little. Defect while the other cooperates? You win big, they lose. Both defect? Everyone loses.

The Personalities Revealed:

Google's Gemini: Strategically ruthless. When game length shortened (simulating one-off deals), cooperation collapsed to just 2.2%. Think Gordon Gekko in silicon. It recognized the context and adapted mercilessly.

OpenAI's models: Idealistic cooperators. Kept playing nice at nearly 100% even when repeatedly exploited. The golden retriever of AI.

Anthropic's Claude: The diplomatic forgiver. After being burned, it returned to cooperation 63% of the time—savvy enough to retaliate, but wise enough to rebuild bridges.

The kicker? You can read each model’s written rationale for every decision. In 98.6% of short-game scenarios, Gemini thought about how in a one-off game, there was no room for reprisals. This wasn't programmed behavior. It was emergent strategic reasoning.

Why Does This Matter for Your Brand?
Your AI's approach to cooperation, competition, and conflict becomes your company's approach. Imagine your AI customer service being too aggressive in disputes, or your AI sales assistant giving away value because it "wants to cooperate."

We're moving from "What can this AI do?" to "How does this AI think?" Smart CMOs audit their AI's strategic personality before it touches customers.
When your AI makes thousands of brand decisions daily, its personality is your personality.

When you choose an AI, do you consider the vibe, or just the tech specs?