Description
HACHuman-led verdict, AI-assisted body
Claude review
Anthropic’s Claude is one of the strongest assistants for long-document analysis, writing and coding. This Claude review covers the 2026 plans, an Ethics Score across six governance dimensions, and a Claude ethical review of the lab that markets itself on safety.
Official link, not a paid placement — independence policy.
How this review works: the Ethics Score rates six dimensions out of ten (60 total) from publicly verifiable evidence and documented conduct — not vendor marketing, not a hands-on lab test. The verdict is the author's own. Pricing changes often; confirm on the vendor's page before buying.
What is Claude?
Claude is a family of models — Opus, Sonnet and Haiku — from Anthropic, used through claude.ai, the API and Claude Code. It handles long-context reading, structured writing, research and software work, with Projects to group files and instructions. Anthropic positions itself as the safety-forward frontier lab, which sets the bar this review holds it to.
Key capabilities
Claude review: 2026 pricing
- Free$0 — daily caps
- Pro$20/mo ($17 annual)
- Max$100/mo (5×) or $200/mo (20×)
- Team$25–$30/seat/mo
- EnterpriseCustom
Claude ethical review — Ethics Score 37/60
Publishes model cards and safety research, but releases selectively and controls what evidence reaches the public.
Paid tiers exclude training by default; consumer opt-out exists (company representation).
Substantial alignment and red-team work relative to peers; key claims remain self-certified.
Reasonable enterprise terms; training-data sourcing is not fully disclosed.
Documented contradictions between safety branding and conduct, including defence-contracting disputes (Verified).
No mass-harm pattern documented; main concern is unverifiable safety self-certification.
Anthropic builds the safety-first brand harder than any peer, which is exactly why this Claude ethical review holds it to that standard. The recurring gap: commitments are voluntary, the company controls what evidence is released, and the strongest safety claims are self-certified rather than independently checkable.
FAQ
Is Claude free?
Yes. Claude has a free tier with daily caps. Pro is $20/month ($17 annual), Max is $100/month (5×) or $200/month (20×), and Team runs $25–$30 per seat (five-seat minimum).
Does Anthropic train on your Claude chats?
Anthropic states it does not train on paid-tier data by default, and Team/Enterprise disable training contractually; consumers can opt out. These are the company’s own representations.
Why does Claude score the way it does here?
Capability, privacy and safety effort score well; corporate conduct is marked down for documented gaps between safety branding and behaviour. The score is an editorial assessment, open to revision as evidence changes.
Reviews
There are no reviews yet.