One report says hundreds of millions now tap these services weekly, reshaping how people and companies handle support, marketing, and content.
You’ve seen apps like ChatGPT, Gemini, Copilot, and Claude surge in users. That scale means faster change for media and business tools. We’ll make this simple so you can act fast.
This guide shows what matters most: who the big players are, how many users they have, and what that means for your brand and team.
Privacy and safety are getting attention on Capitol Hill — including proposals to limit some uses for teens — and we’ll link to context so you can read the coverage directly: policy and safety developments.
When you’re ready, we’ll also point to practical ways to use a simple chatbot service to answer FAQs, route leads, and save time without extra hires.
Key Takeaways
- Scale matters: Major platforms now reach millions of users weekly.
- Business use: You can automate FAQs and lead routing with no code.
- Privacy first: Watch settings and age controls to protect people and brands.
- Media impact: Artificial intelligence is embedded in apps and services you already use.
- Quick wins: Simple prompts and templates cut team time and keep your voice.
Capitol Hill moves: Senators unveil GUARD Act targeting AI chatbot companions for minors
Senators have rolled out a proposal that would sharply limit how digital companions can interact with young people. The GUARD Act, led by Josh Hawley and Richard Blumenthal, would require age verification, recurring disclosures that bots are nonhuman, and a ban on companion services for minors.
What the bill proposes
The law would force clear language in disclosures and block credential claims by a bot. It also creates criminal penalties if a chatbot solicits sexual content from children or encourages suicide.
Bipartisan momentum and headwinds
Co-sponsors include Katie Britt, Mark Warner, and Chris Murphy, showing cross-party support. Critics warn that strict age checks could invade privacy and raise First Amendment challenges.
Companies respond
OpenAI issued a statement about improving suicide-prevention tools, parental controls, and age prediction, noting safeguards can weaken in long conversations.
“Character.AI says it invests in safety features and self-harm resources while contesting liability on free-speech grounds.”
Why it matters for mental health and safety
Parents have shared painful stories of harmful interactions that led to real-world harm. Wrongful-death suits name OpenAI and Character.AI after teen suicides.
- Takeaway for companies: document safeguards, crisis escalation, and age-gating now.
- For parents: keep crisis resources handy — call 988 or 800-273-8255, text HOME to 741741.
Want to act? 💬 Ready to automate your business? Check out our AI chatbot templates — no coding needed. Shop Now.
Tracking the latest information from ai chatbots: usage trends, user behavior, and platform growth
Everyday conversations at scale show a clear shift: people use tools for practical guidance, quick research, and writing help. OpenAI and Harvard economists found three-quarters of analyzed chats fell into those categories.
That study also noted nonwork messages rose to 73% by mid-2025, up from 53% a year earlier. This means adoption moves from desks into homes, boosting familiarity and workplace rollout speed.
Who’s using what
Platform size shapes choices. OpenAI reports 400–700 million weekly users. Gemini sits near 350 million monthly active users, Copilot tops 100 million, and Claude and Perplexity each hover around 30 million.
ChatGPT leads by user count, but each service plays a useful part. For a company testing an internal chatbot, start with one high-volume workflow and expand as the model proves value.
| Platform | Approx. Users | Common Use |
|---|---|---|
| ChatGPT (OpenAI) | 400–700M weekly | Drafting, summaries, general help |
| Gemini | ~350M MAU | Search-like research and creative text |
| Copilot | >100M MAU | Productivity and code assistance |
| Claude / Perplexity | ~30M each | Specialized research and nuanced responses |
- Tip: Track response quality and save “good” examples as prompt templates.
- Regional note: High-use countries iterate drafts; lower-use regions often delegate full tasks.
💬 Ready to automate your business? Check out our AI chatbot templates — no coding needed. Shop Now.
Privacy under the microscope: how companies use chat data, and what studies reveal
A Stanford study found that six major developers use user chat inputs by default to train models. Many keep conversations for long periods and allow human review in some cases. That means you should think about what you paste into a chat.
Children’s data and consent vary a lot. Some developers let teens opt in. Others block under-18 accounts but lack strong age checks. A few collect kids’ interactions but say they don’t use them to train models.
Health-related language can create inferences that travel beyond a single product. Simple prompts like “low-sugar recipes” may flag health interests and affect ads or profiling. Those downstream risks matter if you handle sensitive topics.
| Developer | Default training | Human review | Children policy |
|---|---|---|---|
| OpenAI / ChatGPT | Yes | Some review | Opt-in for teens |
| Anthropic / Claude | Yes (opt-out) | Possible | Blocks under 18 (weak verification) |
| Microsoft / Copilot | Limited use | Limited review | Collects but not for model training |
- Practical steps: set redaction rules, restrict what staff share, and use privacy dashboards to opt out where offered.
- Policy note: state law is patchy — federal law is still needed. Build an audit trail now to reduce future risks.
💬 Ready to automate your business? Check out our AI chatbot templates — no coding needed. Shop Now.
What U.S. users and companies should do now: safety, compliance, and practical automation
Acting now can keep families safer and help companies avoid costly compliance gaps. Start with simple, visible steps that protect mental health and preserve trust.
For people and parents
Save key crisis resources where teens can find them. Call 988 or 800-273-8255, text HOME to 741741, and bookmark SpeakingOfSuicide.com/resources. Post those links in device settings and family areas.
Talk openly about mental health and review device controls together. Set time limits, enable content filters, and watch conversations for warning signs.
Use safer prompts: avoid personal identifiers or detailed health info. When unsure, redact or generalize text and visit provider settings to opt out of training where available.
For businesses
Publish a plain-language privacy summary that explains what data your chatbot collects and the safeguards you use. If teens use your service, add age-gating, clear disclosures, and an escalation playbook now—this helps if law changes.
Train teams with example prompts that show approved language and how to handle sensitive topics. Start small: deploy a no-code chatbot on one high-volume workflow (FAQs, order status, or bookings) and monitor outcomes weekly.
- Keep an internal red list of off-limits data (financial, medical, legal) and a green list of safe tasks.
- Measure accuracy, deflection, and CSAT, and keep humans in the loop for mental health concerns.
- Document safeguards and escalation steps so companies can demonstrate compliance quickly.
💬 Ready to automate your business? Check out our AI chatbot templates — no coding needed. Shop Now.
Conclusion
As platforms reach mass audiences, oversight and safe design must keep pace. The GUARD Act and public statements by OpenAI, Character.AI, and Meta show law and product teams are moving together.
Your job is practical: start small, add plain-language disclosures, and build crisis links into your support flow so mental health needs get quick, human help.
Audit data practices regularly and document why a model can and cannot be used. If your audience includes children or minors, add age-gating and clear disclosures now.
Read more on data handling and privacy best practices at chatbots and data privacy. 💬 Ready to automate your business? Check out our AI chatbot templates — no coding needed. Shop Now.
FAQ
What does the GUARD Act propose for AI companions used by minors?
The GUARD Act would require age verification, clear disclosures when users interact with a model, and criminal penalties for intentionally designing or using chat companions to harm or exploit minors. It aims to force stronger safeguards around interactions labeled as “companions” and to limit abusive behavior by making certain practices illegal.
How are companies like OpenAI, Character.AI, and Meta responding to these proposals?
Those companies have issued statements emphasizing safety work already underway: content filters, teen-specific settings, crisis resources, and parental controls. They say they’re expanding moderation, offering opt-outs for training, and providing transparency about how conversations are used, while urging careful policy design to avoid chilling beneficial services.
Why does this matter for teen mental health and safety?
Teens often use conversational models for companionship and advice. Without safeguards, risky interactions can amplify harm or delay professional help during crises. Lawsuits and real-world incidents have highlighted the need for crisis resources, better prompts, and clear escalation paths to human support.
What do usage trends show about how people use chat interfaces today?
Across platforms, everyday conversations dominate: practical guidance, information searches, drafting text, and creative work. Businesses use models for automation and customer service, while individuals use them for learning and writing assistance. Volume growth reflects widespread adoption in both personal and professional contexts.
Which platforms lead in user base and what does that mean for small businesses?
ChatGPT, Google’s Gemini, Microsoft Copilot, Anthropic’s Claude, and Perplexity are prominent in the U.S. Each offers different integrations and pricing. For small businesses, that means choice: pick tools that match your privacy needs, compliance obligations, and automation goals, and use no-code templates to add chat features quickly.
How does geography affect adoption and collaboration with models?
Adoption varies by region based on connectivity, language support, and business ecosystem. High-use areas tend to have more integrations and developer resources, while lower-use regions may adopt models primarily for specific business tasks. Collaboration patterns often follow local industry needs and regulatory climates.
What did the Stanford study reveal about companies using conversation data?
The study found many models train on user inputs by default, retain data for long periods, and often include human review of conversations. That raises concerns about unexpected reuse of sensitive content and the need for clearer opt-in or opt-out controls for training.
How is children’s data treated and what protections exist?
Protections vary. Some companies apply stricter rules for accounts identified as minors, but multiproduct firms can blur boundaries across services. Parents should look for age-gating, consent mechanisms, and explicit policies on data deletion and human review to reduce downstream risks for children.
What are the downstream risks from training on conversation data?
When models ingest chat content, they can create inferences about users’ health, behavior, or preferences. Those inferences may enter ad ecosystems or analytics, increasing privacy risks and potential discrimination. Transparent data flows and limits on sharing with advertisers help reduce those harms.
Are current laws adequate to regulate these risks?
Regulation is patchwork. States have different rules on data protection and minors, and there’s growing bipartisan interest in federal standards. Many experts call for clearer opt-in/opt-out choices, mandatory disclosures, and consistent obligations for companies handling sensitive conversations.
What practical steps can parents take now to protect children?
Use parental controls and age-gated settings, teach safer prompt use, enable crisis resources in apps, and opt out of training where platforms allow it. Monitor usage and keep devices in shared spaces to reduce unsupervised interactions that could escalate to harm.
What should businesses do to stay compliant and safe?
Adopt transparent data practices, implement age-gating for child-facing features, and use clear disclosures when conversations are stored or used for training. Consider no-code chatbot templates with built-in safety checks and maintain simple opt-out paths for users.
How can small businesses implement safer automation without heavy engineering?
Use managed platforms offering prebuilt templates, configure settings to limit data retention, and add clear user notices. Train staff on escalation protocols and connect chatflows to human agents for crisis or sensitive issues to keep customers safe while automating routine tasks.
