AI & SaaS intelligence

Monitor the AI platforms your product depends on.

API terms change, rate limits shift, and data-handling policies update — and your product breaks or your compliance fails. ConductAtlas tracks the AI ecosystem your stack is built on and archives the evidence.

352+ platforms monitored daily · archived with cryptographic evidence · free to browse, no account.

Monitored AI & SaaS platforms
OpenAI
Anthropic
Google Gemini
AWS
Google Cloud
Hugging Face
Together AI
Perplexity AI
+6 more monitored
352+ platforms monitored daily Archived with cryptographic evidence Governance diffs & version lineage
What you'll monitor

Governance changes affecting the AI services in your stack

Recent changes →
API terms

API & usage terms

Changes to acceptable-use policies, output restrictions, and commercial-use terms that govern how you can ship AI features.

Rate limits

Rate limits & quotas

Shifts in rate limits, tier requirements, and quota structures that can throttle or break your product at scale.

Data handling

Data & privacy policies

Training-data usage, retention, and zero-retention terms that determine whether you stay compliant with your own customers.

Model lifecycle

Model deprecations

Deprecation notices and model sunset timelines that force migrations before your integration stops responding.

Pricing

Pricing & billing terms

Token pricing, minimum commitments, and billing-policy changes that reshape your unit economics overnight.

Compliance

Security & compliance

DPA changes, sub-processor updates, and certification shifts that flow straight into your own SOC 2 and customer contracts.

Why it matters

A vendor policy change can break your product or your compliance

Risk · availability

Integration breakage

A model deprecation or API change can break production with little notice. Migrations under deadline pressure are expensive and risky.

Typical deprecation window: 30–90 days
Risk · compliance

Compliance exposure

Data-handling or sub-processor changes upstream can put you out of compliance with your own customer DPAs without you knowing.

Sub-processor terms change quietly
Risk · scale

Rate-limit surprises

Quota and rate-limit changes can throttle your product during your highest-traffic moments, exactly when you can least afford it.

Limits shift without prominent notice
Risk · margin

Margin erosion

Token pricing and minimum-commitment changes flow directly into your cost of goods. A pricing shift can erase your margin.

Pricing pages update without changelogs
Recent changes

Real changes detected by ConductAtlas

All detected changes →
AWS Bedrock Aug 18, 2026 High
AWS Service Terms
The updated terms establish AWS's authority to suspend and permanently close accounts based on spend limit violations, and introduce new team member content ownership and liability rules that affect how data and intellectual property are governed within shared projects. Organizations using AWS Bedrock should understand that enabling spend limits creates exposure to account suspension and permanent content deletion without a specified reactivation grace period. Project owners become liable for team member conduct and content contributions, which affects internal governance and contract allocation of risk.
AWS Bedrock updated its Service Terms to introduce a new AWS (new) program and restructured billing and project management terms. The update adds new provisions governing project creation, team member access, content ownership for contributed team content, and spend limits with account…
all usersaws bedrock developersteam-based project usersusers with spend limits enabled
View full change record →
Cursor Aug 14, 2026 High
Cursor Terms of Service
The updated terms fundamentally restructure dispute resolution from a private, streamlined arbitration process to mandatory public litigation in Texas courts with strict filing deadlines. Users in other states or countries face significantly increased friction and cost to pursue claims, and claims not filed within 1-2 years are permanently barred. This change affects access to justice and the practical enforceability of user rights under the agreement.
Cursor replaced its arbitration-based dispute resolution system with a litigation-based one governed by Texas law and filed exclusively in Texas courts. Previously, disputes were resolved through binding arbitration administered by the American Arbitration Association; the updated terms require…
all usersUS users
View full change record →
Perplexity AI Aug 14, 2026 High
Perplexity Enterprise Terms
The updated terms substantially expand Perplexity's unilateral authority to modify which services are covered, introduce marketing data sharing outside the data processing agreement, and establish automatic acceptance of changes through continued use. For organizations with vendor governance frameworks, data processing agreements, or privacy representations tied to specific product scope and processor-only data handling, these changes create new compliance review obligations and potential gaps between what the organization has represented to customers or regulators and what Perplexity's terms now permit.
Perplexity AI expanded its Enterprise Terms to cover a broader range of enterprise products and services beyond just Enterprise Pro and Enterprise Max, adding product-specific terms for Comet and Agentic services. The revised terms introduce new provisions governing marketing data sharing with…
enterprise customersbusiness accountsall users
View full change record →
Fireworks AI Jul 8, 2026 High
Fireworks AI Terms of Service
The updated terms establish mandatory arbitration and eliminate class action and jury trial rights for all disputes, including claims that arose before users agreed to the revised terms. This fundamentally alters how disputes are resolved; users can no longer pursue claims in court or as part of a collective action, and must instead proceed through individual arbitration.
Fireworks AI's updated Terms of Service, detected on July 8, 2026, introduces two major dispute resolution mechanisms: mandatory individual arbitration for all disputes and a class action/jury trial waiver. The prior version (which appears to have been unreadable or unavailable in the…
all users
View full change record →
OpenAI May 28, 2026 High
OpenAI Enterprise Privacy
The updated terms establish that workspace administrators, not individual employees, control access to conversations and data retention periods within enterprise accounts. This represents a material shift in data governance authority that affects how enterprise customers must document employee monitoring practices and may create compliance implications under privacy regulations. The expanded retention exception also permits OpenAI to retain deleted conversations longer than previously stated, creating operational discretion that may complicate data deletion commitments.
OpenAI updated its Enterprise Privacy terms on May 28, 2026 to expand workspace admin authority over end user conversations. Previously, only end users could view their own conversations, and end users controlled retention decisions. The updated terms state that workspace admins can now view,…
enterprise customersworkspace administratorsend users in enterprise accounts
View full change record →
How monitoring works

Continuous policy surveillance infrastructure

Capture

Daily snapshots

Automated pipeline captures policy documents daily. Snapshots archived with SHA-256 verification.

Detect

Structural diff

Provision-level diffs identify exactly what changed — not page-level, clause-level detection.

Classify

Severity & impact

Each change classified by severity, direction, affected groups, and operational domain.

Alert

Operational intelligence

Structured alerts with governance analysis, available actions, and compliance implications.

Pricing

Start monitoring your AI dependencies

See full pricing →

Free access to the governance archive. Upgrade for same-day alerts and the vendor risk dashboard across your whole stack.

Stay ahead of the changes

Watch this before it changes again

Follow unlimited companies, monitor the clauses that matter across every platform, and get the full institutional analysis on what each change obligates you to do.

FAQ

Common questions

Which AI and SaaS platforms do you monitor?
We monitor OpenAI, Anthropic, Google Gemini, AWS, Google Cloud, Hugging Face, Together AI, Perplexity AI, Groq, Fireworks AI, Mistral, Replicate, Cursor, and Vercel, among others. Coverage expands continuously and you can request specific platforms.
What kinds of changes will I be alerted to?
API and usage-term revisions, rate-limit and quota changes, data-handling and training-data policy updates, model deprecations, pricing changes, and security/compliance updates such as DPA and sub-processor changes. Each change is classified by severity and operational impact.
Why does this matter for compliance?
When an upstream provider changes how it handles data or who its sub-processors are, that can put you out of compliance with the DPAs you signed with your own customers. ConductAtlas surfaces those upstream changes so you can update your posture before an audit does it for you.
How is this different from watching changelogs myself?
ConductAtlas identifies exactly what changed at the provision level, classifies severity, and archives every version with cryptographic evidence. Most policy and data-handling changes never appear in a public changelog, they land quietly in terms documents most teams never re-read.
Can I monitor platforms beyond AI vendors?
Yes. ConductAtlas monitors 352+ platforms across AI, cloud, payments, and developer infrastructure. You can add any monitored platform to your watchlist.

Know before a vendor change breaks your product.

Governance intelligence for AI and SaaS founders across OpenAI, Anthropic, Gemini, AWS, and the platforms your product depends on. The archive is free to read.

Get the research letter

Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.