Prompt-injection defense
Untrusted content is rendered as data, not commands. Admins put high-risk tools behind human approval, so an injection can't trigger gated actions like moving money or pushing code on its own.
Built so your credentials never touch the AI, every sensitive action waits for approval, and your data never trains a model.
Get Started for FreeChatGPT
ChatGPT in Slack: Search, write, summarize and get work done.
Claude
Anthropic's AI agent for any task — think, write, and code with Claude.
Viktor
Your AI employee in Slack
The audit reports are real, the controls are continuously monitored, and the next audit is always on the calendar.
SOC 2 Type 1
Status
CertifiedCoverage
Independent attestation that our security controls operate as designed. Type II in progress.
Documentation
Report available under NDA.
GDPR
Status
AlignedCoverage
EU data protection requirements met.
Documentation
DPA available on request.
CCPA
Status
CompliantCoverage
California Consumer Privacy Act requirements met.
Documentation
Privacy documentation available.
Slack App Directory
Status
ListedCoverage
OAuth scopes and security posture vetted before shipment through the Slack store.
Documentation
Public App Directory listing.
ISO 27001
Status
In progressCoverage
ISMS controls implementation and evidence collection in progress.
Documentation
Controls overview available today; audit evidence shared after certification.
Here's exactly what Viktor touches — and what he never does.
TLS 1.2+ in transit. AES-256 at rest. Secrets in dedicated vaults.
SAML SSO across Okta (inside Slack), Entra ID, Google Workspace, OneLogin and any SAML 2.0 IdP.
US-hosted by default. EU data residency available on Enterprise contracts.
Admins can disconnect any integration, pause any user, or kill a running task in one click.
Your conversations and files never enter a training set — not ours, not our model providers'.
API keys and tokens are injected at execution time by the tool gateway; the model never sees them.
Money moves, code pushes, and customer emails wait for your explicit approval in Slack.
Skills, integrations, and memory are walled off per workspace. No cross-tenant access.
AI Safety
A backend tool gateway injects your API keys and OAuth tokens at execution time. The AI model itself never sees them.
Not a policy. The architecture.
model-context-preview
Task
Sync HubSpot MQLs
to Salesforce
OAuth token
[Redacted]API key
[Not Available]tool-gateway
Credential
Injected server-side
Vault
AES-256 at rest
Audit
Access logged
AI employees introduce attack surfaces traditional SaaS does not have. Three controls keep the surface small.
Untrusted content is rendered as data, not commands. Admins put high-risk tools behind human approval, so an injection can't trigger gated actions like moving money or pushing code on its own.
Inference runs on OpenAI, Anthropic, and Google. Each is on the public sub-processor list with a no-training agreement for Viktor traffic.
Memory is scoped to your workspace, encrypted at rest, never used to train models, and fully exportable or deletable on request.
Credentials & secrets
Every major tool connects via OAuth with the narrowest scopes that get the job done. No passwords stored.
Where API keys are required, they are stored in a secrets vault, AES-256 at rest, isolated from model context, access-logged, and rotatable.
Admins decide which integrations are connected, who can use them, and at what level. Revoke any integration in one click.

If you believe you have found a security vulnerability in Viktor, we want to hear from you. Email security@viktor.com with enough detail to reproduce the issue: affected URL or component, steps, impact, and any proof-of-concept. Please do not include customer data in the report.
viktor.com, app.viktor.com, api.viktor.com, the Viktor Slack and Microsoft Teams apps, and the Viktor desktop apps. Third-party services we integrate with are out of scope; report those to the vendor.
give us reasonable time to remediate before public disclosure, do not access or modify data that is not yours, do not degrade the service (no DoS, spam, or social engineering of staff or customers), and stop and report immediately if you encounter customer data.
research that follows this policy is authorized, and we will not pursue or support legal action against you for it. If a third party takes action against you for activity conducted in accordance with this policy, we will make it known that you acted in good faith.
we do not run a paid bug bounty today. Meaningful reports receive a thank-you, public credit if wanted, and Viktor credits.
Email security@viktor.com. We acknowledge within 5 business days, provide triage status within 30 days, and fix confirmed critical issues within 90 days. Full policy under Vulnerability disclosure on this page; machine-readable contact at /.well-known/security.txt.
One hire. The output of a team.
Viktor works nights, remembers every decision, and connects to 3,200 tools. Start free with $100 in credits, then $50 a month.