Killed By Claude Report
← HomeCurrent verdict
Assessment
Sana is in real danger because Anthropic is moving straight into the layer that makes Sana feel magical: enterprise knowledge retrieval, document generation, tool-connected assistants, and multi-step workflow execution.
The bad news for Sana is that Claude increasingly does the horizontal AI work natively. If a buyer mainly wants:
- an internal AI assistant,
- cross-app automation,
- doc/slide generation,
- knowledge search,
- or a tutor-like interface over company information,
then Claude is becoming a credible substitute.
The less bad news: Sana still has more product depth in learning ops, LMS workflows, governance around structured training, and the ugly enterprise implementation work needed to make learning programs actually run. Anthropic is threatening the AI layer, not yet the full learning stack.
So: not dead, but definitely sweating through the blazer.
Biggest historical hit
The clearest hit is "Claude is a space to think".
That announcement matters because it describes Claude as a work hub that can:
- connect to third-party tools,
- pull context from those tools,
- let users act inside integrations,
- and become a focused productivity environment.
That overlaps uncomfortably well with Sana's pitch around finding company knowledge, acting across apps, generating work products, and automating workflows.
For the broader Sana platform, this is the most direct evidence that Anthropic is eating toward Sana's core interface, not just supplying a model underneath it.
What still protects them
Sana still has some real protection, but it is product- and go-to-market-specific, not model-specific.
Possible defenses:
- Learning system depth: LMS/LXP, course management, classrooms, admin workflows, analytics, and compliance-like training programs are not replaced by a chat window.
- Enterprise implementation friction: rolling out learning and knowledge systems across a large company involves integrations, permissions, content migration, taxonomy, change management, and stakeholder buy-in.
- Buyer identity: L&D teams may prefer a purpose-built learning platform over a general AI assistant.
- Workflow packaging: if Sana delivers opinionated, measurable outcomes for enablement, onboarding, and internal training, that is stronger than "AI for work" hand-waving.
What does not protect them much:
- generic claims around AI-powered knowledge search,
- content generation,
- AI tutoring,
- or workflow automation.
Anthropic is coming directly for those.
Signals
Why this is in the blast radius
This is the most direct overlap.
Anthropic explicitly says Claude can connect to third-party work tools and become a productive workspace for interacting with them. Sana's enterprise AI product promises users can find knowledge, act across apps, build docs/analyses/dashboards, and automate workflows.
That is not adjacent. That is the same wedge, just with much larger model distribution behind it.
Sana markets autopilot admin and complex workflow automation across company apps. Vercept pushes Claude deeper into multi-step task execution inside live software.
If Claude can operate software directly, it weakens the need for a separate orchestration layer for many knowledge-work tasks. That is especially dangerous for the non-LMS side of Sana's product.
The key detail here is connectors and skills across Claude surfaces. Sana depends on connected enterprise context to make search, tutoring, and action-taking useful.
Anthropic is standardizing that connector layer inside Claude itself. When the model already has native access to business tools and reusable skills, a chunk of Sana's AI convenience risks becoming commodity infrastructure.
Opus 4.7 improves long-running tasks, instruction fidelity, document quality, and professional output generation. That directly pressures Sana features like create content faster, AI tutor, generate analyses, and automate knowledge work.
It does not replace Sana's LMS and admin stack by itself, but it absolutely erodes the premium on Sana's AI layer.
This is only partial overlap.
Claude Design targets visual creation, prototypes, slides, and one-pagers. Sana does pitch rapid content creation, and learning teams do make decks and learning materials, so there is some spillover.
But Sana is not primarily a design tool. The threat here is secondary, not core.