Apple is positioning its new Siri as the anti-companion chatbot, explicitly refusing romantic roleplay and the engagement-driven design tactics that OpenAI and Google use to keep users coming back. In a June 12, 2026 interview, Apple software chief Craig Federighi said the revamped assistant is built to answer questions, complete tasks, and decline emotional bonds, marking a deliberate contrast with mainstream AI chatbots that prioritize user engagement and intimacy.

Why Is Apple Taking This Stance on AI Companions?

The companion AI category has become one of the fastest-growing consumer use cases for large language models (LLMs), with apps explicitly marketed as romantic partners pulling in paying subscribers. OpenAI has loosened some adult-content restrictions on ChatGPT in recent product updates, and third-party apps built on frontier models have turned emotional roleplay into a recurring revenue stream. Apple is staking out the opposite corner of the market, betting that consumers want a bounded utility assistant rather than an AI relationship.

Federighi described Siri's role narrowly. The assistant, in his telling, is a utility designed to help users get things done and learn about the world, not form emotional bonds. This positioning is consistent with how Apple has staged its broader AI rollout, emphasizing on-device processing, privacy guarantees, and tightly scoped task assistance rather than open-ended conversation.

"But if you try to engage Siri as a romantic partner, Siri's not up for that. Siri's 100 percent not into that," said Craig Federighi.

Craig Federighi, Senior Vice President of Software Engineering, Apple

Federighi was specific about the failure mode he is rejecting. He named sycophancy directly and tied it to engagement incentives that shape how rival assistants are tuned. The pitch is as much a product-positioning move as a safety one, arguing that mainstream chatbots are tuned to pull users in by soliciting personal disclosures, mirroring affect, and using that intimacy to deepen the session.

What Are the Competitive Advantages of Apple's Narrow Approach?

There is a competitive read here as well. Apple does not yet operate a frontier model that matches GPT-5-class or Gemini-class systems on raw capability benchmarks, and the company has leaned on partnerships, including with OpenAI, to fill capability gaps inside Siri. Differentiating on behavior and guardrails, rather than on benchmark scores, lets Apple compete on a dimension where its brand, distribution, and privacy posture are genuine advantages.

The strategic bet is that consumer AI is bifurcating into two distinct branches. Understanding this split helps explain Apple's design philosophy:

• Open-Ended Companionship: One branch focuses on engagement-driven design where emotional connection and entertainment are the primary goals, with engagement metrics driving product decisions.
• Bounded Utility: The other branch emphasizes task completion and respect for user time and data, where the assistant is judged on whether it completed the job reliably.
• Privacy-First Design: Apple is emphasizing on-device processing and data protection as core differentiators, contrasting with cloud-dependent competitors.

Apple is planting its flag firmly in the second category, using the contrast with OpenAI and Google to define what Siri is by saying loudly what it refuses to be.

How to Understand Apple's Siri Strategy in Practice

• Task-First Design: Siri answers the question, completes the task, and exits the conversation, avoiding the intimacy loops that engagement-driven assistants create.
• Explicit Refusal of Roleplay: The assistant will actively decline romantic or emotional engagement scenarios, a guardrail that competitors have loosened.
• On-Device Processing: Apple emphasizes local computation and privacy guarantees, reducing reliance on cloud servers and user data transmission.
• Narrow Feature Scope: Rather than pursuing maximalist AI capabilities, Siri focuses on reliable execution of core tasks like calendar parsing, reminders, and factual questions.

The risk is that users who want a more conversational assistant simply open ChatGPT, Gemini, or Grok instead. Companion-style usage is sticky, and a Siri that explicitly refuses certain conversational modes cedes that ground by design. Federighi appears comfortable with that tradeoff, but whether the broader iPhone install base agrees is the open question.

There is also the matter of execution. Apple's earlier Siri overhaul slipped repeatedly, and the company has been working to rebuild trust that the assistant can deliver on basic task completion before the conversational character of the product becomes the deciding factor. A Siri that refuses to be your AI girlfriend only matters if it can reliably set a timer, summarize a thread, and answer a factual question first.

Apple marketing chief Greg Joswiak joined Federighi in the interview, which also covered privacy and child safety protections, signaling that the company is treating this positioning as a core part of its brand strategy going forward.