Introduction – AI is No Longer a Gadget
In three years, ChatGPT has transitioned from a technological curiosity to a daily tool. According to the latest report from OpenAI, nearly 700 million people use it weekly by mid-2025. Users are more diverse than ever: in 2025, 52% are women, compared to only 37% at the beginning of 2024.
Generative AI no longer belongs to tech enthusiasts; it is becoming part of everyday life.
Meanwhile, Anthropic, creator of the AI Claude, publishes its own figures. They show a clear divergence:
- ChatGPT mainly serves as a consultant,
- Claude is used as an automaton capable of executing entire tasks.
These two visions reveal a major shift: we no longer speak of "AI" in the singular, but of specialized uses. The era of the one-size-fits-all model is behind us.
ChatGPT: The Virtual Consultant of Everyday Life
The OpenAI report shows that more than half of conversations involve requests for information or advice.
ChatGPT acts as a versatile assistant:
- helping to write or improve a text (emails, posts, reports);
- providing practical information;
- organizing ideas or preparing written work.
More technical uses (programming, artistic creation) remain marginal. ChatGPT is primarily a facilitator of daily tasks.
Another notable fact: in 2025, 73% of conversations are personal, compared to an equal private/professional use in 2024. The tool is used as much to organize life as to work.
In the B2B world, it is also entering purchasing processes: 90% of organizations report using it to search for or compare products.
Claude: The Automaton for Businesses
Claude, on the other hand, takes a different path. According to Anthropic, 77% of its business uses involve task automation. Where ChatGPT assists, Claude does it instead.
Among the tasks assigned:
- code generation and complex calculations;
- document processing and reports;
- email management and other repetitive processes.
📈 In eight months, the share of uses where the user entrusts a complete task to the AI has risen from 27% to 39%.
Companies trust Claude to execute from start to finish.
This choice fuels the debate on employment. Dario Amodei, CEO of Anthropic, predicts that AI could eliminate half of entry-level office jobs within five years. Whether one agrees with this vision or not, one thing is clear: AI is already a player in the workforce, not a futuristic gadget.
Increasing Specialization of AIs
These reports indicate the end of the fantasy of the "universal Swiss army knife." Like business software (Word ≠ Excel ≠ accounting software), AIs are specializing:
- ChatGPT excels as a generalist companion, advisor, and writer;
- Claude establishes itself as an automaton for delegating entire tasks;
- tomorrow, industry-specific AIs will emerge: legal, medical, financial, creative…
each optimized for its function.
The consequence is that AI is no longer an interchangeable tool, and choosing the right AI becomes strategic.
Issues of Trust and Control
This segmentation does not eliminate risks. Three points consistently arise:
-
Confidentiality – Public chatbots like ChatGPT do not guarantee data security.
➤ Sending sensitive information may violate confidentiality. -
Reliability – Hallucinations exist.
➤ American lawyers have already been sanctioned for citing case law invented by AI. -
Responsibility – In case of error, it is always the user (or the signing professional) who remains responsible.
In short, AI does not replace human vigilance.
⚖️ Focus on Law: LegalTech and Generalist AI, a Complementary Duo
In the legal sector, this trend towards specialization takes an obvious form: LegalTech.
✅ These tailored solutions:
- integrate confidentiality guarantees;
- rely on solid doctrinal and jurisprudential bases;
- establish themselves as secure co-pilots.
Should we then dismiss generalist AIs? Not necessarily. ChatGPT and similar tools retain their usefulness for:
- producing a first draft of text;
- structuring an idea;
- stimulating reflection.
The future of law will not be a duel between generalist AI and LegalTech, but a reasoned combination:
- LegalTech for the core of the profession, where security and precision are critical;
- public AI as a creative assistant and accelerator upstream.
Conclusion – Taming Rather Than Enduring
The reports from OpenAI and Anthropic confirm that AI is becoming democratized and specialized. ChatGPT, Claude, and tomorrow's other AIs illustrate a world where the tool will no longer be "the same for everyone," but chosen based on its use. For lawyers and other professionals, the challenge is no longer if AI should be used, but how and with which tool.
And above all, it is important not to approach this new ecosystem alone, as beyond the technical aspect, it is a cultural change and a shift in working methods.
When well-placed, it frees up time and energy. When misused, it undermines quality and trust.
The future, therefore, is about learning to choose and combine. Generative AI, each to their own!
📚 Sources:
- OpenAI, How people are using ChatGPT, Sept. 2025
- Anthropic, AI Usage Index, Sept. 2025
- Washington Post, OpenAI study reveals who uses ChatGPT, Sept. 2025
- Business Insider, Studies of ChatGPT Versus Claude, 2025
- Entrepreneur, Businesses Are Using AI to Automate Work, 2025
- Hendy, ChatGPT for Lawyers: Opportunities and Best Practices, 2023