Terms of Service

1.1 The website agentbeam.com and its sub-sites (including the documentation, learning, blog, dictionary, CVE database, attacks tracker, releases and public analytics pages), the cloud dashboard at app.agentbeam.com and its APIs, the Agentbeam command-line tools and local runtime recorder (the “CLI”), Studio, the Beam Sentinel and Agentbeam workspace browser extensions, the App monitoring service and SDK, the Beam MCP server, the team library, integrations and alert workflows, the in-product assistant, self-hosted and enterprise packages, and every other product, feature, update or service we make available under the Agentbeam name (together, the “Platform”) are owned and operated by AISOLO Technologies Private Limited [CIN: U62099MH2023PTC403351] (“Agentbeam”, “we”, “us”, “our” or “Company”), having its registered office at 1003, Kamdhenu Commerz, Sector 14, Kharghar 410210, Navi Mumbai, Maharashtra, India.

1.2 These terms and conditions (“Terms”) form a legally binding agreement between the Company and each person or entity that accesses or uses the Platform (“you”, “your”, “User”). Where an organization signs up, connects devices, or an individual uses the Platform on its behalf, “you” includes that organization (the “Customer”).

1.3 You accept these Terms by creating an account, installing or running any part of the Platform, connecting a device, browser or application to a workspace, clicking to accept, signing an order, or otherwise using the Platform. If you do not agree, do not use the Platform. These Terms incorporate our Privacy Policy, our End User License Agreement, our GDPR Compliance page and our Data Rights page, each as updated from time to time.

1.4 Agentbeam provides a security and observability layer for AI agents, AI-assisted work and the applications that use them (collectively, the “Service”). Descriptions of features, on the Platform, in documentation or in marketing, are summaries. The documentation describes how a feature is intended to work and what it does not do; features differ by plan, role, release, operating system and deployment.

2.1 Separately signed agreements. If a separate written agreement signed by an authorized representative of both parties exists, it governs to the extent it addresses the same subject matter, and these Terms continue to apply to everything it does not address. An online click-through, a purchase order, a vendor security questionnaire, or terms attached to an order or procurement portal do not replace or modify these Terms unless we have signed them in writing.

2.2 Order of precedence. If documents conflict, the following order applies, highest first: (i) a signed agreement under Section 2.1, and within it the order it states; (ii) any order form, quote or plan description we have accepted, as to commercial terms (scope, seats, fees, term) only; (iii) any signed data processing addendum, as to the processing of personal data only; (iv) these Terms; (v) the End User License Agreement, as to the installed software only; (vi) the Privacy Policy, GDPR Compliance page and Data Rights page; and (vii) our documentation and other published materials, which are for information and are not contractual commitments unless an order or signed agreement incorporates them.

2.3 Free and individual use. If you use the Platform without paying, including the free tier, the open source CLI, Beam Sentinel, the learning, dictionary, CVE, attacks and blog content, or a trial, these Terms are the complete agreement and apply as written.

2.4 Affiliates and resellers. If you buy through a reseller, marketplace or channel partner, these Terms still govern your use of the Platform. Your commercial arrangements with that party are separate and do not give you rights against us beyond these Terms, and we are not liable for the reseller's acts or omissions.

2.5 Interpretation. Headings are for convenience only. “Including” means “including without limitation”. A reference to a law includes that law as amended. Capitalized terms not defined in a section have the meaning in Section 3.

4.1 Eligibility. You must be at least 18 years old, or the age of majority in your jurisdiction, to create an account. The Platform is not directed at anyone under 16 (see the Privacy Policy). If you use the Platform on behalf of an organization, you represent that you have authority to bind it to these Terms, and that organization is responsible for your acts and omissions.

4.2 Registration. You must provide accurate, current information and keep it up to date. We may refuse, block or remove accounts, including those using disposable or spam email addresses, or that we reasonably believe are created to abuse the Platform.

4.3 Credentials and tokens. You are responsible for safeguarding passwords, sessions, invitation links, device credentials, pairing tokens, App monitoring tokens, integration URLs and keys, signing secrets, and any API key or third-party key you place in the Platform, and for all activity under them. Treat invitation links and tokens as credentials. Notify us immediately at [email protected] of any suspected compromise or unauthorized use.

4.4 Roles and administration. A Workspace has roles (such as Owner, Administrator, Security manager, Auditor, Billing and Member) that determine what each person can see and do. Owners and administrators control membership, policy, devices, integrations, and collection settings. You are responsible for assigning the least privilege necessary, for the acts of the people you give a role, and for promptly removing access and revoking devices when people leave. We may rely on any instruction that appears to come from a person with the required role in your Workspace.

4.5 Owner privacy setting. Data about a Workspace owner's own devices, activity, sessions, findings, skills, browser activity, usage and audit entries is hidden from other roles by default until an owner chooses to share it. This is a visibility control inside your Workspace, not a data-deletion or retention control, and does not limit our access as described in the Privacy Policy, our obligations under law, or what your own policies require.

4.6 One person, one account. Do not share an Account between individuals. Seat counts include members and unexpired pending invitations.

The Platform is modular. Depending on your plan, role and deployment, it includes some or all of the following. Each is subject to these Terms, the Documentation's statement of limitations, and any additional terms shown when you enable it.

We may add, change, rename, limit or discontinue features. If we materially reduce paid functionality during a paid term, we will give notice and, where we cannot offer a reasonably equivalent replacement, refund the prepaid fee for the affected period after the change takes effect.

6.1 Right to use. Subject to these Terms and payment of applicable fees, we grant you a limited, non-exclusive, non-transferable, non-sublicensable, revocable right during your term to access and use the Platform for your own internal business or personal purposes, within the scope of your plan, seats and allowances. Installed software is additionally licensed under the EULA.

6.2 Open-source CLI. The Agentbeam CLI (the @agent-beam/beam package, published at github.com/whyashthakker/beam-cli) is open source under the GNU Affero General Public License v3.0 only. Its license file, not these Terms, governs your rights to copy, modify and redistribute the CLI's source code, and nothing in these Terms adds restrictions to those rights. Everything else is proprietary and governed by these Terms: the cloud dashboard and APIs, the website, the browser extensions, App monitoring service, the Beam MCP server, enterprise companions, hosted services, and the Agentbeam name and marks. Using the CLI with our hosted services is also subject to these Terms.

6.3 Reservation of rights. All rights not expressly granted are reserved. No rights are granted by implication, estoppel or otherwise.

The Platform observes software, devices, browsers, people and traffic. You are solely responsible for deciding what to observe and for doing so lawfully. In particular, you:

• Represent that you own, or are legally authorized to administer, every device, browser profile, account, application, server, domain and log you connect or point the Platform at
• Are the party that decides to monitor your personnel, contractors and customers, and must give every notice, obtain every consent, consult every works council or representative body, and complete every assessment that Applicable Law requires before you connect their devices, browsers or activity to a Workspace, enable text collection, or send them alerts
• Must configure collection, redaction, retention, role access and enforcement to match your own legal, contractual and policy obligations, and must not use the Platform to covertly surveil individuals in breach of Applicable Law
• Are responsible for the content of the prompts, source code, files, logs and other material your agents, users and applications process, and for having the right to submit it to the Platform
• Must not submit data for which Applicable Law or contract requires a safeguard the Platform does not provide, such as regulated health data, payment card data, or classified or export-controlled data, unless we have agreed in writing that the Platform is suitable for it
• Are responsible for the security of your devices, networks, cloud accounts, repositories, credentials, logs and backups, including the local storage where the recorder writes
• Are responsible for reviewing flagged activity and for every action your agents, users and applications take, whether or not the Platform flagged, paused or blocked it
• Must test enforcement and policy changes on representative devices before wide rollout, and keep your own backups, version control and recovery plans
• Must keep the CLI, extensions and SDK reasonably up to date to receive rule, policy and security fixes

Rules that look for personal data, secrets or sensitive patterns are aids, not legal compliance tools. Using the Platform does not make you compliant with any law, standard or certification, and our records are not a substitute for your own compliance program.

You agree not to, and not to permit anyone else to:

• Use the Platform to violate any Applicable Law, contract, or third party's rights, including intellectual property, privacy and employment rights
• Attempt to circumvent, disable, tamper with, or interfere with security-related features, hooks, policy, licensing, seat limits, rate limits or allowances of the Platform, except as the Documentation expressly allows an administrator to configure
• Reverse engineer, decompile or disassemble any component that is not open source, or attempt to extract source code, models, rules or the proprietary rule sets, except to the extent Applicable Law expressly permits despite this restriction
• Resell, sublicense, white-label, or offer the Platform as a service to third parties, or use it to provide a managed service, without our prior written consent
• Use the Platform, or data or access obtained through it, to build or benchmark a directly competing product or to train a competing model, or publish benchmark results of non-public components without our written consent
• Introduce malware, probe, scan or test the vulnerability of our systems or other users' systems without written permission (see our security contact for coordinated disclosure), or attempt to gain unauthorized access to systems, accounts or data
• Send traffic or data intended to overload, degrade or abuse the Platform, its ingestion endpoints, APIs, assistant or alerting, or exceed documented limits or allowances through automation or multiple accounts
• Use App monitoring or any scanner against an application, host or network you do not own or have written authority to test and monitor
• Use alerts, webhooks or integrations to send spam, unlawful content, or to harass; or direct integrations at destinations that do not belong to you
• Publish to the team library any content that contains malware, secrets, personal data you have no right to share, infringing material, or instructions designed to subvert an agent or reviewer
• Misrepresent your identity or affiliation, create accounts with disposable or spam addresses, or interfere with another user's use of the Platform
• Use the Platform in connection with weapons development, unlawful surveillance, or any activity that causes or is likely to cause death, injury or serious harm

We may investigate suspected violations and may suspend, throttle or terminate access, remove content, revoke tokens or block addresses, with notice where reasonably practicable and without liability. We may report unlawful activity to authorities.

9.1 Plans. Agentbeam is available on a free tier for individual use and on paid team and enterprise plans that add cloud sync, fleet visibility, App monitoring capacity, administrative controls and other features. Current plans, seat counts and allowances are shown on the Platform or in your Order before purchase. Plan features, allowances and limits, including monthly App monitoring and LLM-review allowances, rate limits and storage or history limits, can change prospectively with notice as described in Section 29.

9.2 Billing. Subscription fees are billed in advance for the selected billing period through our payment processor, Stripe, and renew automatically for successive periods of the same length unless cancelled before the renewal date. You authorize us and our processor to charge the payment method on file, including for seat additions made during a term, for which we may charge pro rata. Seat changes, payment methods and cancellation are managed in the billing portal. We do not store full card details.

9.3 Taxes. Fees are exclusive of applicable taxes, duties and levies, including GST, VAT and sales tax. You must provide accurate billing details, including legal entity name and any tax identification number, and are responsible for all taxes other than taxes on our income. If withholding is required by law, you will gross up the payment so that we receive the full invoiced amount, unless a valid exemption certificate is delivered before payment.

9.4 Late and failed payment. If payment fails or is overdue, we may retry the payment method, charge interest on overdue amounts at 1.5% per month or the maximum lawful rate if lower, recover reasonable collection costs, and after notice suspend or downgrade the Workspace. Suspension does not relieve you of fees owed.

9.5 Disputed charges. Notify us of a billing dispute in good faith within 30 days of the invoice date at [email protected]. You must pay undisputed amounts when due.

9.6 Refunds. Except where required by law or expressly stated in your Order or at purchase, fees are non-refundable and no credit is given for unused seats, allowances or periods.

9.7 Price changes. We will give at least 30 days' notice of a price increase before it applies to your next renewal. If you do not agree, cancel before the renewal.

9.8 Trials and promotions. Free trials, credits and promotions may be changed or withdrawn, are provided without warranty, and convert to paid plans only as disclosed when you start them.

9.9 Self-hosted licenses. A self-hosted dashboard uses an installation-bound, signed license that states edition, expiry and seat count and is validated offline. An invalid or expired license stops new configuration, invitations, enrollment and ingestion; reading and exporting history, reviewing risks and revoking access remain available. Invoicing, support and renewal for self-hosted installations follow your license agreement or Order. You are the operator and are responsible for infrastructure, security, backups, retention and upgrades.

10.1 Ownership. As between you and us, you own Customer Data. Nothing in these Terms transfers ownership of Customer Data to us.

10.2 License to us. You grant us and our sub-processors a worldwide, non-exclusive, royalty-free license during the term (and afterwards for the retention periods in the Privacy Policy and as needed to comply with law and resolve disputes) to host, copy, transmit, process, analyze, display and back up Customer Data solely to provide, maintain, secure, support and improve the Service for you, to enforce these Terms, to prevent fraud and abuse, and to comply with law. We do not sell Customer Data, and we do not use it to train third-party foundation models.

10.3 Usage Data and improvement. We may collect and use Usage Data, and may use Customer Data in de-identified and aggregated form that does not identify you, your Workspace or any individual, to operate, secure and improve the Platform, to develop detection rules and threat intelligence, and to publish aggregate statistics (for example the public analytics and attack trends). We own Usage Data and the resulting rules and insights. Customer-selected destinations (such as your own LLM key) are excluded.

10.4 Feedback. If you give us suggestions, ideas or feedback, you grant us a perpetual, irrevocable, worldwide, royalty-free right to use it without restriction or attribution.

10.5 Our property. We own the Platform, including the scanning heuristics, rules, rule sets, threat-intelligence content, detection logic, product design, documentation, the Agentbeam and Beam names and logos, and all intellectual property in them, except for components expressly licensed as open source or content owned by third parties.

10.6 Data backup and export. Local data remains under your control on your device. Cloud history can be exported using the Platform's export features (for example the audit log CSV export, which is bounded). Exports are copies and may not be complete. Preserve evidence you need before disconnecting devices, uninstalling, removing members or deleting a Workspace.

11.1 Privacy Policy. Our Privacy Policy describes what we collect and why. Our GDPR Compliance and Data Rights pages describe rights and how to exercise them.

11.2 Roles. For personal data contained in Customer Data (for example the names, email addresses, devices and activity of your members, and the IP addresses, user agents and request details of visitors to applications you monitor), you are the controller (or business) and we act as your processor (or service provider), processing on your documented instructions as set out in these Terms, your configuration of the Platform, and any signed data processing addendum. For account, billing, security, website and product analytics data we process for our own purposes, we are an independent controller.

11.3 Data processing addendum. A Data Processing Addendum, including Standard Contractual Clauses and the UK Addendum where required, is available on request at [email protected]. Where you sign one, it prevails as to personal data under Section 2.2. Until then, the processor commitments in the Privacy Policy, GDPR page and this Section apply. We will use the sub-processors listed on our GDPR page, may change them on reasonable notice, and impose data protection obligations on them that are no less protective than those in this Section.

11.4 Your obligations. You are responsible for the lawfulness of the personal data you submit or cause to be collected, for giving notices to, and obtaining consents from, the individuals concerned (Section 7), for responding to their requests, and for configuring the Platform to minimize the personal data collected. We will assist you with data subject requests and regulatory enquiries to the extent the Platform does not already let you do so yourself, and may charge reasonable fees for excessive requests.

11.5 Security incident. We will notify the affected Customer without undue delay after we become aware of a personal data breach affecting Customer Data we process. Notification is not an admission of fault or liability.

12.1 Our measures. We use reasonable technical and organizational measures designed to protect Customer Data in our systems, including encryption in transit, access controls and least-privilege principles, hashed storage of App monitoring tokens, and audit logging of administrative changes. No method of transmission or storage is completely secure. Agentbeam is an early-stage product and does not hold itself out as SOC 2 or ISO 27001 certified unless and until we say so in writing.

12.2 Local-first responsibility. Activity recorded locally stays on your device by default. You are responsible for securing the device and storage location, for backup, and for retention or deletion in line with your policies. We have no access to, and cannot recover, local data that has not been synced.

12.3 Remote actions you authorize. Where you enable them, authorized roles can ask connected devices, through the Workspace, to remove a skill, send the text of one SKILL.md file (redacted and size-limited), update the CLI to a named release, or honor an address pause or deny list. The request is carried out by software on the device at its next poll, only on items in the device's own inventory, and is recorded in the audit log. You authorize us to transmit these requests on your behalf, you are responsible for who in your Workspace holds the permission, and you accept that a requested action can fail, be delayed, or be ignored (for example by older software, an offline device or a device setting that opts out) and that deleting or updating software can disrupt work. Test before relying on it. Environment settings documented for a device can disable these actions.

12.4 Network blocking and address pauses. Pausing or denying a sender address, and any firewall or deny-list function, can block legitimate users who share the address (for example behind a proxy or office network). It acts only where your application or web server consumes the result. You decide when to use it.

12.5 Vulnerability disclosure. Report suspected vulnerabilities to [email protected]. Do not access data that is not yours, degrade the service, or disclose a vulnerability before we have had a reasonable opportunity to address it. We will not pursue good-faith research that follows these rules.

13.1 Integrations you configure. The Platform integrates with Third-Party Services that you connect or configure, such as AI providers and coding assistants, MCP servers, IDEs, chat and paging tools (Slack, Microsoft Teams, Discord, Google Chat, PagerDuty), webhooks, email, cloud and edge providers, package registries, and your own models or decision services. They are governed by their own terms and privacy practices. When you enable one, we send data to it on your instruction, you are responsible for that transfer and for any charges, and we are not responsible for how the third party handles it, for its availability, security, output or changes.

13.2 Your own keys. Where you supply an API key (for example for LLM review of findings, action judging through a general model, or a TypeSafe Jev key), the data goes to that provider under your agreement with it and may incur charges. Redaction before sending is heuristic and cannot remove every sensitive value. Inspect and minimize inputs before enabling remote checks.

13.3 Intelligence and reference sources. Vulnerability, exploit and threat information is compiled from third-party and public sources (for example public vulnerability catalogs and advisories). We do not control and do not warrant their accuracy, completeness or timeliness.

13.4 Service providers we use. We use providers listed on our GDPR Compliance page to operate the Platform, such as hosting, payments, email, analytics and AI providers.

13.5 Links. The Platform may link to third-party websites or resources. We are not responsible for their content, accuracy or practices.

14.1 What they are. The Platform includes features that use machine-learning or large-language models, including the in-dashboard assistant, LLM review of App monitoring findings, tips and recommendations, and action judging you may enable. The assistant is given your question, documentation and, for permitted roles, counts and labels from your Workspace (not finding details, device hostnames or member emails). Questions typed into the assistant may be logged in redacted, truncated form to improve documentation. Do not enter secrets or personal data into the assistant.

14.2 Output can be wrong. AI output is probabilistic and may be inaccurate, incomplete, out of date, biased or manipulated by adversarial input, including input written by an attacker. A “benign” or “allow” verdict is not a safety guarantee. Review AI-assisted output before relying on it. Do not use it as the sole basis for decisions with legal or similarly significant effects on individuals; flagged items are for human review.

14.3 Your agents. The Platform monitors AI agents you run but does not control them. We are not responsible for any action, output, cost, data exposure or damage caused by an AI agent, coding assistant, model, MCP server, skill or package, whether or not the Platform observed, flagged, paused or blocked it.

14.4 Training. We do not use Customer Data to train third-party foundation models.

Beta Features are provided for evaluation “as is”, may contain errors, may change or be withdrawn at any time, may not be covered by support or any availability commitment, and you use them at your own risk. Features marked “coming soon”, “roadmap” or similar describe intent, not a commitment, and you should not purchase in reliance on them. Our liability for Beta Features is limited to the maximum extent permitted by law, and in all cases to the lower of the amount paid for them and USD 100.

16.1 Content you publish to a team library (skills, instruction files, commands, subagents, rules) is made available to your Workspace according to your sharing and review settings. You represent you have the right to share it, that it does not infringe or contain secrets or malware, and you grant us the license in Section 10.2 to host and scan it. Authors and administrators are responsible for what they publish and approve.

16.2 Scanning and review are heuristic and an “approved” version is not a warranty that it is safe, correct or fit for any purpose. Installing library content on a device runs under your authority and is your responsibility.

16.3 We may remove content that violates these Terms or Applicable Law. If you submit other content (for example to support, in a community channel, or as a report), you give us a license to use it to respond, operate and improve the Platform.

The blog, learning course, dictionary, CVE database, attacks tracker, response guides, comparisons, benchmarks, release and stress reports, public analytics and similar content are provided for general information. They may contain summaries of events or claims reported by third parties, some of which are marked unverified, and analysis that reflects our judgment at the time of writing. They are not legal, security, financial or professional advice and may be incomplete or out of date. Comparisons are based on information available to us and represent our opinion; product names belong to their owners and no affiliation or endorsement is implied. Public analytics are aggregate and are not a statement about any specific person or organization. If you believe something is inaccurate, contact [email protected] and we will review it.

Each party (the “Receiving Party”) agrees to protect the other's (the “Disclosing Party”) non-public information that is marked confidential or that a reasonable person would understand to be confidential (“Confidential Information”), such as non-public product roadmap, pricing, security details, and, for you, Customer Data, using no less care than it uses for its own confidential information of a similar nature and at least reasonable care. The Receiving Party will use it only to exercise rights or perform obligations under these Terms and will not disclose it except to employees, contractors, advisers and sub-processors who need to know and are bound by equivalent duties, or as required by law (with prior notice where lawful).

Confidential Information excludes information that is or becomes public without breach, was known to or independently developed by the Receiving Party without use of the Disclosing Party's information, or was rightfully received from a third party. These obligations survive for three years after termination, and for as long as Customer Data remains in our possession.

19.1 Availability. We aim for high availability of the cloud dashboard and ingestion but do not guarantee uninterrupted or error-free operation. Scheduled maintenance, emergency repairs, provider outages and attacks may cause downtime, delay or data loss. Ingestion is best-effort: forwarding batches are held in memory, are not a durable retry queue, and events may be delayed, dropped or absent from the dashboard. Local history on the device remains the record of what was captured. No service-level agreement applies unless stated in a signed agreement or Order.

19.2 Support. Support is provided by email at [email protected] on a commercially reasonable basis, with response targets only where stated in your Order. Free-tier support is best-effort.

19.3 Force majeure. Neither party is liable for delay or failure to perform (other than payment) caused by circumstances beyond its reasonable control, including natural disasters, epidemics, war, terrorism, civil unrest, labor disputes, government action, internet, power, cloud or telecommunications failures, failures of upstream providers, and malicious attacks.

20.1 Agentbeam is a detective and, where enabled, a preventive control. It is one layer, not a complete security program. Without limitation:

• Detection is heuristic pattern matching, not semantic malware analysis. It will miss attacks, abuse and exposures for which it has no rule, and some legitimate activity will match
• Redaction removes recognized credential formats only and cannot classify all confidential information; source code, prose and personal data may remain
• Scanners, package checks and dependency checks reflect known indicators and public data at the time; a clean result is not proof of safety
• Hook-, extension- and SDK-based capture depends on a client that invokes the hook and honors its response; it does not intercept every process, network path, native app, API call or MCP transport, and it is not a kernel sandbox
• Browser coverage is limited to supported sites and interactions; attachments in binary formats may not be inspected; provider changes can break coverage
• A recorded command or request shows what an agent proposed, not what the operating system completed
• Operating-system coverage and optional enterprise monitors vary and must be validated on your own deployment
• Alerts, findings, roll-ups, severities, scores and verdicts may be false positives, false negatives, delayed, duplicated or missing
• Records are not a lossless transcript and are not WORM storage; administrators of a database can alter them in self-hosted installs

THE PLATFORM, INCLUDING ALL SOFTWARE, CONTENT, AI OUTPUT, RULES, SCORES AND REPORTS, IS PROVIDED “AS IS” AND “AS AVAILABLE”. TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW, WE AND OUR LICENSORS AND SUPPLIERS DISCLAIM ALL WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, INCLUDING MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, TITLE, NON-INFRINGEMENT, ACCURACY, QUIET ENJOYMENT, AND ANY WARRANTY ARISING FROM COURSE OF DEALING OR USAGE OF TRADE. WE DO NOT WARRANT THAT THE PLATFORM WILL BE UNINTERRUPTED, TIMELY, SECURE OR ERROR-FREE; THAT IT WILL DETECT, PREVENT, BLOCK, REMEDIATE OR INVESTIGATE ANY THREAT, ATTACK, VULNERABILITY, DATA LEAK OR MISUSE; THAT DATA WILL BE COMPLETE OR ACCURATE; THAT ANY DEFECT WILL BE CORRECTED; OR THAT USE OF THE PLATFORM WILL ACHIEVE COMPLIANCE WITH ANY LAW, STANDARD OR CERTIFICATION, OR WILL PREVENT ANY LOSS.

THE PLATFORM IS NOT DESIGNED OR LICENSED FOR USE IN ENVIRONMENTS WHERE FAILURE COULD LEAD TO DEATH, PERSONAL INJURY OR SEVERE PHYSICAL OR ENVIRONMENTAL DAMAGE. NO ORAL OR WRITTEN ADVICE FROM US CREATES A WARRANTY. SOME JURISDICTIONS DO NOT ALLOW CERTAIN DISCLAIMERS, SO SOME OF THE ABOVE MAY NOT APPLY TO YOU; IN THAT CASE THEY APPLY ONLY TO THE EXTENT PERMITTED.

22.1 Excluded damages. TO THE MAXIMUM EXTENT PERMITTED BY LAW, NEITHER AGENTBEAM NOR ITS AFFILIATES, OFFICERS, DIRECTORS, EMPLOYEES, CONTRACTORS, LICENSORS, SUPPLIERS OR SUB-PROCESSORS (THE “AGENTBEAM PARTIES”) WILL BE LIABLE UNDER OR IN CONNECTION WITH THESE TERMS OR THE PLATFORM, WHETHER IN CONTRACT, TORT (INCLUDING NEGLIGENCE), STRICT LIABILITY, BREACH OF STATUTORY DUTY, MISREPRESENTATION, RESTITUTION OR ANY OTHER THEORY, FOR ANY: (A) INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, EXEMPLARY OR PUNITIVE DAMAGES; (B) LOSS OF PROFITS, REVENUE, SALES, BUSINESS, CONTRACTS, GOODWILL OR REPUTATION; (C) LOSS OR CORRUPTION OF DATA, OR COSTS OF RECOVERING OR RECREATING DATA; (D) BUSINESS INTERRUPTION OR DOWNTIME; (E) COST OF PROCURING SUBSTITUTE GOODS OR SERVICES; OR (F) LOSS ARISING FROM AN INCIDENT, BREACH, ATTACK, EXFILTRATION, DESTRUCTIVE ACTION, VULNERABILITY OR AGENT ACTION THAT THE PLATFORM DID NOT DETECT, FLAG, PAUSE OR BLOCK, OR DETECTED OR BLOCKED LATE, INCORRECTLY OR INCOMPLETELY, OR FROM A FALSE POSITIVE OR FALSE NEGATIVE, IN EACH CASE EVEN IF ADVISED OF THE POSSIBILITY AND EVEN IF A REMEDY FAILS OF ITS ESSENTIAL PURPOSE.

22.2 Liability cap. TO THE MAXIMUM EXTENT PERMITTED BY LAW, THE AGENTBEAM PARTIES' TOTAL AGGREGATE LIABILITY FOR ALL CLAIMS ARISING OUT OF OR RELATING TO THESE TERMS AND THE PLATFORM WILL NOT EXCEED THE AMOUNT YOU PAID US FOR THE PLATFORM IN THE 12 MONTHS BEFORE THE FIRST EVENT GIVING RISE TO LIABILITY, OR USD 100 IF YOU HAVE NOT PAID US (INCLUDING ON THE FREE TIER). MULTIPLE CLAIMS DO NOT ENLARGE THE CAP, WHICH IS A SINGLE AGGREGATE CAP FOR ALL CLAIMS AND ALL PARTIES TO WHICH IT APPLIES.

22.3 Allocation of risk. The parties agree that the disclaimers, exclusions and caps in these Terms reflect a reasonable allocation of risk, are an essential basis of the bargain, and that the fees reflect that allocation. The Platform observes and aids, it does not insure; we cannot and do not assume responsibility for the security of your systems, the conduct of your agents, people or applications, or the consequences of the decisions you make with Platform output.

22.4 Remote and destructive actions. We are not liable for any loss arising from an action you or a member of your Workspace requests, schedules or authorizes through the Platform, including software removal, updates, address pauses, deny lists, enforcement decisions or other device commands, or from the failure, delay or partial execution of such an action.

22.5 Third parties. We are not liable for Third-Party Services, for acts or omissions of third parties, for actions of your members, agents or applications, or for loss arising from your failure to follow the Documentation, keep software updated, secure credentials, or maintain backups.

22.6 Matters not limited. Nothing in these Terms excludes or limits liability that cannot be excluded or limited under Applicable Law, including liability for fraud or fraudulent misrepresentation, or for death or personal injury caused by negligence, where such law applies. For clarity, the exclusions and cap in this Section do not apply to your payment obligations, or to your liability for breach of Section 7 or Section 8, your indemnity in Section 23, or infringement or misappropriation of our intellectual property.

22.7 Time limit. To the extent permitted by law, any claim against an Agentbeam Party must be brought within one year after the cause of action arises, or it is permanently barred.

22.8 Individuals and consumers. If you are a consumer and mandatory consumer-protection law gives you rights that this Section would reduce, those rights are not affected.

22.9 Beneficiaries. The Agentbeam Parties are intended beneficiaries of this Section and Section 21 and may enforce them.

23.1 By you. You will defend, indemnify and hold harmless the Agentbeam Parties from and against all claims, demands, proceedings, investigations, losses, liabilities, damages, fines, penalties, settlements, costs and expenses (including reasonable legal fees) arising from or relating to: (a) Customer Data, or your submission, collection or use of it, including any allegation that you lacked the right, notice or consent to collect or submit it; (b) your or your Authorized Users' violation of these Terms or Applicable Law; (c) your monitoring of personnel, devices, browsers, applications, networks or traffic; (d) actions you or your members request or authorize through the Platform, including remote and destructive actions; (e) your agents, applications, code, skills, packages and content; (f) a dispute between you and any of your members, personnel, customers or end users; or (g) your use of Third-Party Services.

23.2 Process. We will notify you promptly of a claim (delay relieves you only to the extent you are prejudiced), you will control the defense and settlement with counsel reasonably acceptable to us, you will not settle a claim imposing liability or admissions on us without our consent, and we may participate with our own counsel at our cost.

23.3 By us. We provide no indemnity under these Terms. A signed agreement under Section 2.1 may provide a limited indemnity, which is then the sole remedy for the matter it covers.

24.1 Term. These Terms apply from first acceptance until terminated. A paid subscription runs for the period in your Order and renews under Section 9.

24.2 By you. You may stop using the Platform at any time, uninstall the software, disconnect devices, and cancel a paid subscription from your account or billing settings; cancellation takes effect at the end of the current billing period and fees already paid are not refunded except as provided in Section 9.6.

24.3 By us. We may suspend or terminate access immediately on notice (or, where urgent, without prior notice and informing you promptly after) if you breach these Terms, fail to pay, are insolvent, if required by law or a provider, to protect the security, integrity or availability of the Platform or other users, or if we stop offering the Platform or a feature (in which case we will refund prepaid fees for the unused period of the discontinued paid service). We may terminate free access at any time.

24.4 Effect. On termination, your right to use the Platform ends, you must stop using it and delete or uninstall software, outstanding fees become due, tokens and device credentials are revoked, and we will delete Customer Data as set out in the Privacy Policy (including a recovery period of up to 45 days after account deletion), subject to legal retention obligations. Local data on your devices remains yours; software removal may be needed. After termination, you may request an export within the recovery period; we have no obligation to retain data beyond it.

24.5 Survival. Sections that by their nature should survive termination do, including Sections 2, 3, 7 (as to prior conduct), 8, 9 (payment obligations), 10, 11 (to the extent of data we retain), 13, 14, 17, 18, 20–23, 25, 26 and 28–30.

25.1 Governing law. These Terms and any dispute or claim arising out of or in connection with them or the Platform (including non-contractual disputes) are governed by the laws of India, without regard to conflict of law principles.

25.2 Informal resolution. Before starting a proceeding, the party raising a dispute must give written notice describing it, and the parties will try in good faith to resolve it within 30 days. This does not prevent either party from seeking urgent injunctive relief.

25.3 Courts. Subject to Section 25.2, the courts at Mumbai, Maharashtra have exclusive jurisdiction over any dispute arising from these Terms, except where mandatory consumer-protection law gives you the right to bring a claim in your local jurisdiction. We may seek injunctive relief for infringement or misuse of our intellectual property or confidential information in any competent court.

25.4 Individual claims. To the extent permitted by law, claims may be brought only in an individual capacity and not as a plaintiff or class member in any purported class or representative proceeding.

You will comply with export control, economic sanctions and anti-corruption laws that apply to your use of the Platform, including those of India and the United States. You represent that you are not located in, a resident of, or controlled by a person in a comprehensively embargoed country, and are not on any restricted-party list, and will not use or export the Platform to any such person or place, or for any prohibited end use. Neither party will offer or accept any bribe or improper payment in connection with these Terms.

We may give notice by posting on the Platform, in-product message, or email to the address on your Account, to your Workspace owners, or to the security or billing contact you set. Notices to us must be sent in writing to [email protected] with a copy to our registered office. Email notices are treated as received on the next business day after sending. You are responsible for keeping your contact details current.

We will not use your name or logo publicly without your consent, which may be given by email. You may not make public statements implying our endorsement. Neither party may describe the other's security posture publicly without consent, except for accurate statements required by law.

We may update these Terms from time to time, including to reflect new features, Applicable Law or business practices. We will post the updated Terms with a new “Last updated” date and, for material changes, notify registered account holders at least 30 days before they take effect where practicable (or immediately where required by law or to address security). Continued use after the effective date constitutes acceptance. If you do not agree, stop using the Platform and cancel before the change takes effect. Changes do not apply retroactively to disputes arising before the change, and cannot override a signed agreement under Section 2.1.

30.1 Entire agreement. These Terms (with the documents they incorporate) are the entire agreement between you and us for the Platform and supersede prior understandings, subject to Section 2. You have not relied on any statement not set out in them.

30.2 Severability and waiver. If a provision is held unenforceable, it will be enforced to the maximum extent permissible and the rest continue in full force. Our failure to enforce a provision is not a waiver of it.

30.3 Assignment. You may not assign or transfer these Terms without our written consent. We may assign them to an affiliate or in connection with a merger, acquisition, reorganization or sale of assets, and may subcontract performance.

30.4 Relationship. The parties are independent contractors. There is no partnership, agency, fiduciary or employment relationship. There are no third-party beneficiaries except as stated in Section 22.9.

30.5 Electronic form and language. These Terms may be accepted and notices given electronically, and electronic records are admissible. The English version controls over any translation.

30.6 Government and regulated users. If you are a government body or regulated entity, you remain responsible for any approvals your use requires; additional terms may be set out in a signed agreement.

30.7 Contact. Contact us at [email protected] for legal matters, [email protected] for general support, or [email protected] for security matters. AISOLO Technologies Private Limited, 1003, Kamdhenu Commerz, Sector 14, Kharghar 410210, Navi Mumbai, Maharashtra, India.