AI Companion Watch — 2026-04-01
Microsoft 365 Copilot is making headlines today with its new "Critique" and "Council" modes, which run Claude and GPT side by side on the same task — a striking sign of the multi-model era arriving in earnest. Meanwhile, Google's Gemini has quietly rolled out a notable quality-of-life update: users can now import their full chat history and memories directly from ChatGPT and Claude. Reviewers are also weighing in on the broader companion app landscape, with Nomi, Character.AI, and Candy AI each earning praise in different categories.
AI Companion Watch — 2026-04-01
New Releases
No confirmed new AI companion app launches were reported in the past 24 hours. The most recent coverage focuses on major updates to existing platforms (see below).
Major Updates
Microsoft 365 Copilot — Critique & Council Modes
- What changed: Microsoft has added two new modes to Microsoft 365 Copilot — Critique and Council. In Critique mode, GPT drafts a response and Claude fact-checks and critiques it. In Council mode, multiple AI models (including both OpenAI and Anthropic models) work simultaneously on the same query, with their outputs blended or compared.
- Impact: This is a significant shift for enterprise users. Rather than betting on a single model, Microsoft is now letting competing models cross-check each other's work inside a single workflow. Early-access customers began receiving the Copilot Cowork rollout as of March 30. Accuracy and depth on complex research tasks are the stated goals.
Google Gemini — ChatGPT & Claude Import Tools (Full Rollout)
- What changed: Google has now fully launched switching tools inside Gemini that allow users to import memory summaries, personal preferences, and full chat history from both ChatGPT and Claude. The feature was confirmed live as of March 31, 2026.
- Impact: This significantly lowers the barrier to switching AI assistants. Users who have spent months training ChatGPT or Claude with personal context no longer have to start from scratch. This is a direct competitive move aimed at accelerating Gemini adoption.
User Verdict
Independent reviewer site RoboRhythms published a hands-on roundup of 8 AI companion apps (dated 1 day ago), concluding:
"Nomi has the best memory, Candy AI wins on visuals, Character AI is the best free pick."
The verdict aligns with a broader pattern in companion app reviews: no single app dominates every category, and users are increasingly choosing based on their primary use case — emotional depth, visual interaction, or cost.
Character.AI's scale continues to stand out in market data: the platform has reported 233 million users as of April 2026, with 15 million mobile monthly active users, according to CompanionGuide.AI.
Comparison Corner: Multi-Model Workflows vs. Single-Model Assistants
This week's most interesting dimension is how Microsoft, Google, and Anthropic are positioning their AI assistants — specifically whether they lean into multi-model collaboration or double down on a single best-in-class model.
| Microsoft Copilot | Google Gemini | Claude (Anthropic) | |
|---|---|---|---|
| Approach | Multi-model: GPT + Claude side by side | Single model, but importing rivals' data | Single model; also powering Copilot's Critique layer |
| Key move this week | Critique & Council modes launched | Full ChatGPT/Claude import rollout | Acting as fact-checker inside Microsoft's ecosystem |
| Target user | Enterprise / complex research | Users switching from ChatGPT or Claude | Developers + enterprise via API |
| Openness | High — uses competitors' models | High — imports competitors' data | Neutral — benefits from Copilot deal |
Takeaway: Microsoft is making the boldest architectural bet, treating AI models as interchangeable tools rather than locked-in vendors. Google is competing on portability and lock-in reduction. Claude, meanwhile, is winning partnerships precisely because of its reputation for careful, critical reasoning — now literally the "fact-checker" role in Copilot.
This content was collected, curated, and summarized entirely by AI — including how and what to gather. It may contain inaccuracies. Crew does not guarantee the accuracy of any information presented here. Always verify facts on your own before acting on them. Crew assumes no legal liability for any consequences arising from reliance on this content.
Create your own signal
Describe what you want to know, and AI will curate it for you automatically.
Create SignalPowered by
