Privacy Policy

This policy explains how Efham.ai handles personal data when you use our website, member platform, courses, community, messages, AI learning assistant, image and video studio, payments, and related notifications.

Efham.ai is the service provider referred to in this policy. We do not state an unverified legal entity name or business address. If you use a third-party service connected to the platform, that provider's policy may also apply.

We collect name, email, a hashed password, phone number and country code, age, interface language, country of residence, and nationality, plus optional profile details such as city, company, job title, bio, social links, and membership-question answers.

For email/password registration, we may send a one-time email verification code. With Google sign-in, we receive the provider identifier, verified email, name, and avatar from Google, then ask you to complete required profile details and accept the Terms and Privacy Policy.

We record operational and security evidence such as IP address, session, device and browser information, signup and login timestamps, and may infer the network country at signup, latest login, and latest observation. We use a local MaxMind GeoIP database to map an IP to an approximate country; we do not request precise GPS location.

This network evidence is separate from your declared residence and nationality. We do not automatically overwrite either with the IP country; we may compare signals to reduce fraud, protect accounts, and apply appropriate payment pricing.

We process plan, billing cycle, amount, currency, payment status and dates, provider references, and subscription, cancellation, and renewal details. Egypt pricing is shown in EGP and international pricing in USD or the currency shown at checkout, based on residence and related verification signals.

Card payments and recurring subscriptions are handled by Stripe, and payments in Egypt by debit card, ValU or Fawry may be handled by XPay or Fawry, so Efham.ai does not store full card details. Manual Instapay payments may include an uploaded payment screenshot and remain pending for review, approval, or rejection. We retain transaction and review records for support, accounting, and fraud prevention.

We collect course enrollment, progress, watched episodes, watch time, quiz results and attempts, uploads, and feature or quota usage. This supports learning continuity, achievement measurement, badges, and content improvement.

When you use the Nagaty assistant, we process your prompts, conversation context, attachments, and usage signals, and may send necessary content to the AI provider Anthropic/Claude to generate a response. MongoDB, Redis, and Qdrant may support operational storage, sessions, semantic search, and retrieval context. Avoid entering highly sensitive data that is not needed for the answer.

When you use the image and video studio, we send the prompt you write and any images you upload with it, together with the settings you chose, to the generation provider Higgsfield to produce the image or clip. Your name and email are not sent with the request. Under Higgsfield's terms, it may use what we send it, and what it generates from it, to improve its models; it does not offer us a way to exclude member content from that. So do not upload pictures of people who have not agreed to it, and do not put personal or sensitive details in a prompt that the generation does not need.

We keep a copy of every generated image or clip in Efham.ai's own storage, with its prompt, settings, and cost, and show it to you alone in your work history. No other member sees it unless you publish it yourself. We do not use your prompts or uploaded images to train any model of our own.

When AI assignment review is switched on and you submit an assignment, an AI evaluator may assess it against a marking guide prepared from the assignment brief and its lessons and approved by the team. It reads the text of your submission and the files and images attached to it (for example the text of a PDF or Word document); it does not open links you include, which only a person on the team sees. It may then approve the assignment, or ask for changes with feedback and links to the lessons that can help. To do this we send the submission and the marking guide to one of the AI providers listed in section 10 (OpenAI, Anthropic, or Google). For some assignments the team reviews the evaluation before you receive it. A person on the team can review any decision made this way and change or reverse it; if an approval is reversed, the points it gave are removed and any certificate that depends on the assignment waits until it is approved again.

We process posts, comments, likes, mentions, media, images, videos, GIFs, reports, and moderation decisions. Content posted to community or public-profile areas may be visible to other members according to the service's settings and features, together with your name and avatar.

Direct messages and voice notes are intended for conversation participants, but they are not end-to-end encrypted. We may access them on a limited basis when needed to operate the service, investigate a report or abuse, or comply with a valid legal request. GIPHY may provide GIF search results when that feature is enabled.

Community posts and comments are read automatically to help the team notice what needs a reply, support, or moderation: an unanswered question, a problem using the platform, a complaint, a suggestion, content that breaks the rules, or an achievement worth congratulating. To do this we send the text of the post or comment, together with the author's subscription tier, how long they have been a member, and whether they have already received a reply, to the classification provider TypeSafe (the Jev model), which returns probabilities only, generates no text, and does not train on what we send. Your name and email are not sent with the request. We may then ask the Nagaty assistant (Anthropic/Claude) to draft a reply for the team to review before posting; these drafts are not posted automatically. The results are visible to platform administration only, never to any member, and no automated decision about your account or content is taken from them; the AI community assistants and automated moderation described below work differently.

We also record which community posts you have read. A post counts as read once it has stayed visible on your screen for a few seconds, not merely because it scrolled past. For each post you read we keep the first and most recent read, how many times you returned to it, and whether your subscription was paid at that time. No other member sees this; it is visible to platform administration only, and we use it to measure and improve content reach.

This measurement is first-party and does not pass through external analytics tools, so it is not covered by the analytics-cookie refusal described in section 8. Reads by signed-out visitors are not linked to any person and are counted only as a total. We hold no reading records at all for posts published before this feature was enabled.

Efham.ai may use AI community assistants: accounts that answer questions and help in the community under their own names. When they are switched on, to decide whether an assistant should reply, which one, and how, and to check content against the community rules, we send new posts, comments and replies, including comments on the community posts for lessons, to the AI providers OpenAI, Anthropic and Google (see section 10). We send the whole conversation, meaning the post and the comments and replies around it with the names of the people who wrote them, together with the images, PDFs and other files attached to it. The AI models use it to decide whether a reply is needed, to check the content against the rules, and to draft the reply. Depending on the topic, the reply is either posted automatically under the assistant's name or saved for the team to approve first. The assistants do not answer direct messages.

So that an answer fits you, the assistant also sees some private context about the member who wrote the post or comment: their public profile (name, bio, badges and any partnership label), their membership (plan, whether it is paid, and when they joined), their learning (courses enrolled in, lessons completed, and the goals they set in My goals), their moderation history on Efham.ai (reports about their content, warnings, automated actions and posting mutes), and a private AI-generated summary of their interests, struggles and goals (see the analysis engine paragraph below). This context is private to the AI: the assistant is instructed never to mention or reveal it, we check replies and block any that repeat private details such as your plan or your warnings, and it is never shown to other members. It does not include your email address, phone number, payment or billing history, or support conversations.

An assistant may look things up on the web to answer a question, using the AI provider's search tool. For this we send the provider the question and part of the conversation around it (with the names shown in the community), but not your account details such as your email, phone number or plan; the provider's search tool then runs the searches. Answers that use the web end with links to their sources.

Automated moderation (when switched on): when an AI moderation assistant finds that a post, comment or reply breaks the community rules, it may, depending on how serious and how clear the breach is, post a polite public reminder of the rules, send you a private warning, hide the content while the team reviews it, or remove it. If you break the rules repeatedly within a short period (currently three confirmed breaches within 30 days), you may be muted for 24 hours: during that time you cannot publish posts or comments, but you can still read the community and message the platform owner. Each time, we tell you privately, by an in-app notification, a notification on your device, and a message from the moderation assistant's account, which rule was involved and how to appeal. The team can see every automated action and can reverse it, and appeals are decided by a person.

Analysis engine (when switched on): from time to time the platform owner runs an AI analysis of community posts, comments and replies (each read as a whole conversation), of direct messages only in conversations the platform owner takes part in, and of assignment submissions (their text, files and images). It reads no other direct messages. We use it to build the assistants' knowledge base of common questions and answers, with personal details removed (anything drawn from direct messages becomes general questions and answers, with no names and no link back to the conversation); to give the team insights such as frequent questions, points of confusion in lessons, complaints and suggestions; and to write the private summary of a member's interests, struggles and goals described above. The assistants use this summary and platform administration can see it; it is never shown to other members. For this we send that content to the AI providers listed in section 10.

We process points, levels, streaks, badges, leaderboards, and founding-member status, plus referral codes, referrer relationships, commissions, and affiliate payout requests and details as needed to run referral and affiliate programs.

We may collect NPS scores, surveys, feedback, support requests, and notification and marketing preferences. We use these to understand satisfaction, improve the service, and communicate according to your choices.

When our comeback-reminder feature is switched on, to remind you to come back and keep learning, if you have learned on Efham.ai at least once, we compute once a day, from your activity on Efham.ai only, an internal estimate of how likely you are to come back and learn the next day and within the next week and, if you are a paying member close to your renewal date, how likely you are to renew. It draws on things such as the days you learned, the lessons, articles, and quizzes you completed and your quiz answers, video playback problems, community replies and direct messages you received, your posts and comments, your use of the AI assistant, your streak, points, and level, how you have recently opened Efham.ai, your plan and renewal date, and your latest NPS rating. From the same activity we work out what tends to bring you back (for example new lessons, replies waiting for you, your streak, or progress in a course you started) and the hour at which you usually learn, in your timezone. We use the estimate of coming back and this simple profile to decide whether to remind you, when, on which channel, and with what message. If you ignore several reminders in a row, we send them less often.

These estimates and this profile are internal. They are never shown to other members; only platform administration can see them, to understand which members may be drifting away from learning. They do not change your price, your access, or anything you are entitled to, and no decision with a legal or similarly significant effect on you is made from them automatically.

To measure whether reminders actually help, a randomly chosen group of paying members (currently about 5%) receives no comeback reminders on any channel. Whether you are in this group is decided once, when you first become a paying member, and stays the same, except that a member in it who stops paying leaves it; free members are not placed in it. We still compute the estimate for members of this group so the two can be compared. No other email or notification is affected, such as new-lesson announcements or replies to your posts.

You can switch comeback reminders off at any time: turn off “Learning reminders” under Settings → Notifications, on the web or in the app. We send reminder emails only to paying members who have turned on marketing email, which you can turn off in the same place or with the unsubscribe link in every reminder email. Free members get reminders only as push and in-app notifications, never by email. You can also turn off Efham.ai notifications in your browser or device settings.

We use cookies and local storage needed for sessions, language, security, and interface preferences, including a cookie named efham_consent that remembers your analytics choice for both efham.ai and app.efham.ai. Meta Pixel or Google Ads are used only when configured and enabled, and Meta Pixel and the Google Ads tag load only after you select Accept.

If you select Accept, efham.ai and app.efham.ai also load the OpenAI Ads pixel, OpenAI's advertising measurement tool: it tells OpenAI which of our pages you visit, and when a new account is created and when a new subscription is bought, so we can measure our ads on ChatGPT, and like any web request it carries technical details such as your IP address and browser type. It is not loaded before you accept or after you select Reject. The pixel's automatic advanced matching is switched on: it reads the email address, phone number and name typed into forms on our pages, such as the sign-up form, and turns them in your browser into SHA-256 hashes (scrambled codes that do not show the original data) before sending them to OpenAI, so OpenAI can match these results to people who saw or clicked our ads. When you are signed in, the pixel also sends OpenAI SHA-256 hashes of your account email, phone number, name and account identifier, together with your country and, if you added one, your city, for the same matching. When you pay for a subscription, we also save with your payment the ad-click and browser references this pixel keeps in your browser (__oppref and __obref), so OpenAI can match the purchase to our ads when it is confirmed later. The OpenAI pixel also sets a cookie named __oaiq_consent that records your choice; if you select Reject, the pixel deletes __oppref and __obref and only __oaiq_consent stays, to remember that you refused OpenAI tracking so nothing is sent.

When you create an account on our website, or buy a new subscription by card, through XPay (debit card, ValU or Fawry), through Fawry, or through an Instapay transfer we approve, and you have selected Accept on the web, our servers also send OpenAI, through its Conversions API, a record of that event: for a new account, the time it was created and a coded reference derived from your account that cannot be reversed; for a purchase, the amount and currency actually paid, the plan, and a payment reference; a SHA-256 hash (a scrambled code that does not show the original data) of your account email, phone number, name and account identifier; your country, and your city if you added one to your profile; and, where available, the browser and ad-click references the OpenAI pixel keeps (__obref and __oppref) and the IP address and browser user agent of your request. We never send your email, phone number or name themselves, we do not send renewals, plan changes or AI credit purchases, and nothing is sent if you selected Reject or never answered the cookie question on the web. We use this to measure our ads on ChatGPT.

We use PostHog for product, conversion, and campaign analytics. Before you make a choice, and after you select Reject, measurement on efham.ai and app.efham.ai is anonymous and cookieless: we record the page visited, the referring site and campaign (UTM) parameters, browser and device type, and an approximate country derived from the IP address. Instead of storing an identifier on your device, PostHog computes a hash from the IP address and browser user agent on its servers that changes every day. Apart from the cookies that remember your choice, nothing is stored on your device for analytics, and this measurement is not linked to your account.

If you select Accept, PostHog sets a first-party analytics cookie shared by efham.ai and app.efham.ai, so your visits to both sites are recognised as coming from the same browser. If you sign in or create an account on app.efham.ai, these analytics are linked to your account identifier and account email, together with basic account attributes such as plan, country, and interface language, for product analytics. Your choice applies to both sites.

If you select Accept and arrive from a Google ad, we keep the ad's click identifier (gclid, gbraid or wbraid) in a cookie named efham_adclick on efham.ai and app.efham.ai, and copy it to your account when you sign in or create an account. When this reporting is switched on, our server tells Google Ads when you create an account on our website, when you apply for a scholarship, when you pay for a new subscription by card, through XPay or through Fawry, and when we approve a subscription you paid by Instapay, so we can see which of our ads work. We send the event time and a reference number (for a new account, a coded reference derived from your account that cannot be reversed; for an application, its reference; for a payment, the payment reference), the amount and currency for a payment (converted to US dollars if we report amounts that way), the click identifier if the click was within the last 90 days, and a SHA-256 hash of your account email and, if you added one, your phone number. A hash is a one-way code: Google can match it to an account it already knows, but we never send the email or phone number itself. We do not send your name, plan or IP address, and we tell Google not to use this data to personalise ads. We do this only for members who selected Accept on the web.

We do this only with the consent you give by selecting Accept in the cookie banner. To withdraw it, select Reject: reopen the banner with the “Change cookie choice” link in the site footer, or from Settings → Privacy & Data in the app. Reject deletes the efham_adclick cookie at once; the next time you open app.efham.ai while signed in, we also clear the click identifier from your account, and nothing that happens after that (a payment, a scholarship application) is reported. Withdrawing does not undo events already reported.

If you have selected Accept and are signed in, and Meta Pixel's advanced matching option is switched on, the pixel also sends Meta your account email, phone number, country and account identifier, each turned into a SHA-256 hash inside your browser before it is sent, so Meta can match the visit to an ad.

If you select Accept and arrive from a Meta ad, the efham_adclick cookie also keeps Meta's click identifier (fbclid). When this feature is switched on, for members who selected Accept on the web, our servers send Meta, through its Conversions API, a record of these account events: creating an account on our website; a subscription or purchase, by card, XPay, Fawry or Instapay (the amount and currency actually paid, the plan, and a payment reference); a subscription renewal; submitting an Instapay transfer for review (amount and currency); applying for a scholarship (which scholarship, by its reference number — never what you wrote in the application); and learning milestones (completing your first lesson and completing a course). Each record carries the event name and time and a reference for the event (for a new account or a learning milestone, a coded reference derived from your account that cannot be reversed); a SHA-256 hash of your account email, phone number, country and account identifier; the Meta cookie identifiers (_fbp and _fbc), and the click identifier (fbclid) if you arrived from a Meta ad; and, where available, the IP address and browser user agent of the request. We never send your email or phone number themselves, your name, lesson content, messages or posts. We use this to measure and improve our Meta ads.

To match our ad results to events that happen without you visiting, such as a renewal, we keep the IP address, browser user agent and Meta cookie identifiers (_fbp and _fbc) from your signed-in visits after you select Accept, for up to 30 days. We delete them sooner if you select Reject (the next time you open app.efham.ai while signed in) or delete your account.

We keep ad click identifiers (Google's gclid, gbraid or wbraid, and Meta's fbclid) for up to 90 days after the ad click: the efham_adclick cookie expires then, and we delete the copy saved on your account. We delete them sooner if you select Reject, as described above, or delete your account.

Our servers send Meta nothing about anyone who selected Reject or has not yet answered the cookie banner. Withdrawing your consent by selecting Reject, as described above, stops any further sends but does not undo those already made. The Efham.ai mobile app does not include any Meta tracking code (SDK); your activity in the app is included in these records only if you selected Accept on the web.

These records go from our servers, not from your browser, so they are sent only after you select Accept on the web, and only while you have not selected Reject since — the same choice that controls the pixels. When you create an account, we first save on the account the cookie choice your browser made, and decide from it. We may also stop loading the Meta Pixel, the Google Ads tag and the OpenAI Ads pixel in browsers altogether and measure our ads only through these server records; what is sent, and when, stays as described above.

Our analytics never include the content of posts, comments, messages, AI prompts, or search text, and we do not use session recordings. Page addresses are reduced to the path and campaign parameters before they are sent.

Our servers record account and billing events, such as subscriptions, payments, and referral earnings, and send them to PostHog against your account identifier for billing analytics, regardless of your cookie choice; no cookies are involved in this.

In the Efham.ai mobile app, usage analytics are linked to your account when you are signed in, and you can turn them off in the app's Settings.

Our analytics processor is PostHog, Inc., and this data is stored in the United States (PostHog US cloud). To change your choice on the web, use the “Change cookie choice” link in the site footer, or Settings → Privacy & Data in the app, to bring the consent banner back, or contact us at the address in section 15.

We use Resend for verification codes, service mail, permitted campaigns, and audience synchronization. We record delivery, bounce, and complaint status and suppress invalid or complaining addresses. You can opt out of marketing email; necessary transactional, security, and account messages may continue.

We may use Web Push for notifications and Socket.IO/realtime web connections for presence, messaging, and updates. Web Push requires browser permission, which you can revoke in browser or account settings.

In the Efham.ai mobile app, notifications are delivered through Expo's push service (Expo, Inc.), which passes them to Apple Push Notification service on Apple devices and, when enabled, to Google Firebase Cloud Messaging on Android. For this we store, for each device on which you allow notifications, a push token together with the platform (iOS or Android), device model, app version, and the app's language, and the notification's title and text pass through these services to reach your device. You can turn the app's notifications off in the app's Settings or in your device settings.

When our comeback-reminder feature is switched on and you open Efham.ai while signed in, on the web or in the app, we record how you arrived, for example from an email, a push notification, an in-app notification, the home-screen icon, a community link, a search engine, or directly, together with the first page you landed on (without its link parameters), the platform (web, the web app installed to your home screen, iOS, or Android), and the time and your local hour and weekday. We also record your timezone as reported by your device or browser and, on the web, whether you have allowed, blocked, or not yet answered browser notifications. We do not keep the address of the site that referred you, only the type of source. Links in our own emails and notifications carry two tracking parameters, src (the channel) and nid (an identifier for that message), so we know which message, if any, brought you back. We also send the type of source and the platform to PostHog in the way described for analytics in section 8.

For each comeback reminder we keep a log of when it was planned and sent and on which channel, whether it was delivered, opened, or clicked, and whether you came back or completed a lesson within 24 hours afterwards. For reminder emails, delivery, opens, and clicks are reported to us by Resend, which detects opens and clicks with a small invisible image in the email and links that pass through its servers; for push notifications we record when the push service accepted the notification, not when it appeared on your screen; for in-app notifications we record when you opened them; and a click is recorded when you arrive through the message's link.

We share data as needed with processors and operational providers, including Google OAuth for authentication; Stripe, XPay and Fawry for payments; Resend for email; Anthropic/Claude for the AI assistant and for drafting team replies; TypeSafe (the Jev model) for automatically triaging community posts and comments to find what needs the team's attention or moderation (see section 6); Higgsfield for image and video generation (see section 5 on its use of content to improve its models); PostHog, Inc. for analytics, with data stored in the United States (see section 8); Sentry for error reporting; Cloudflare Turnstile for bot protection; MaxMind for network-country estimation; Linode hosting and Object Storage; MongoDB, Redis, and Qdrant data layers; GIPHY for GIF search when enabled; Meta Pixel, Google Ads, or the OpenAI Ads pixel when enabled (see section 8), which load in the browser only after you accept cookies; Google Ads (via Google's Data Manager API), when enabled, to measure new accounts, scholarship applications and subscription payments (card, XPay, Fawry and approved Instapay) of members who accepted cookies (see section 8); Meta Platforms (via Meta's Conversions API), when enabled, to measure new accounts, subscriptions, payments, Instapay transfers, scholarship applications and learning milestones of members who accepted cookies (see section 8); and OpenAI (via its Conversions API), when enabled, to measure new accounts and new subscriptions of members who accepted cookies (see section 8).

To run the AI community assistants, automated moderation, the analysis engine and assignment evaluation (see sections 5 and 6), when they are switched on, we also share the content described there with OpenAI, Anthropic and Google (Gemini), as processors that run the AI models on our behalf. We use their developer and business (API) services, and we do not use your content to train any model of our own.

We may disclose limited data to advisers or authorities when needed to protect rights or safety, comply with law, or support a legitimate corporate transaction with appropriate safeguards. We do not sell personal data for money.

For the mobile app's push notifications we also use Expo (Expo, Inc.) and, through it, Apple Push Notification service and, when enabled, Google Firebase Cloud Messaging (see section 9).

We use data to deliver the service and perform our subscription agreement, verify identity, operate learning, community, payments, and support, personalize the experience, measure performance, prevent fraud and abuse, maintain security and reliability, and comply with legal obligations.

Depending on applicable law, we rely on service or contract necessity, consent where required, legitimate operational interests not overridden by your rights, and legal compliance. You may withdraw consent for future processing without affecting prior lawful processing. Reporting your new account, scholarship applications and purchases to Google Ads (section 8) relies on your consent alone. Sending your account events to Meta and to OpenAI, and Meta Pixel advanced matching (section 8), also rely on your consent alone.

Our providers may process data in countries other than yours. We use safeguards available and proportionate under applicable law and provider contracts, although data-protection rules vary by country.

We retain data while an account is active or as needed for service delivery, security, dispute resolution, and financial or legal obligations. Periods vary by record type. After deletion, eligible data is erased or anonymized, while limited copies may remain temporarily in backups, fraud-prevention logs, or required financial records until securely rotated.

Community post-reading records (section 6) have a fixed period: we keep them linked to your name for 180 days from your most recent read, after which they are deleted automatically. Only per-post totals remain, not linked to any individual.

Some records of the AI community assistants, the analysis engine and AI assignment review (sections 5 and 6) have fixed periods: the log of each post, comment or reply the assistants read and what they decided about it, including a moderation assistant's assessment of it, is deleted automatically 400 days after it was created; the analysis engine's reports and content suggestions, and the new assistant topics the AI suggests, 400 days after they were created; a member's private summary and the other analysis results, such as knowledge-base entries, 400 days after they were last updated; and each AI assignment evaluation, including drafts that were never sent, 400 days after it was created. The record of the automatic reading of each community post or comment (section 6), including what was sent, the result and any draft reply, is likewise deleted 400 days after it was created. Your thumbs up or down on an assistant's reply is deleted 400 days after you last changed it, and each list of earlier assignment submissions sent together for AI review, 400 days after it was created. The review you received stays with your assignment submission.

Records of automated moderation (the action taken and its reasons, and any reversal or appeal and its outcome) are kept to operate and audit the service: so the team can check and reverse decisions, handle appeals, and recognise repeated breaches of the rules. Like the other records in this section, they are kept while your account is active or as long as they are needed for those purposes, and after account deletion they are handled as described above. Knowledge-base entries built by the analysis engine contain no personal details, and a member's private summary is updated each time the analysis runs again.

Images and clips generated in the studio (section 5) are kept with their prompt and settings for as long as your account exists, because they are your record and your receipt for what was paid. You can ask at any time for all or some of them to be deleted by emailing the address below, and all of them are deleted when you delete your account from settings. Deleting them removes them from our storage and stops any further use of them by us, and stops any new content going to Higgsfield; it does not recall what has already been sent to Higgsfield, which is governed by Higgsfield's own terms.

Some comeback-reminder records (sections 7 and 9) have fixed periods: the record of each time you opened Efham.ai is deleted automatically 400 days after that visit, the daily record of your video playback quality 120 days after the last viewing recorded that day, and the log of each reminder 400 days after it was created.

We use technical and organizational controls including transport encryption, password hashing, access controls, verification, rate limits, logs, and monitoring. Sentry may capture error and technical context for diagnostics. No service is 100% secure, so protect your credentials and report suspicious activity.

Subject to applicable law, you may request access, a downloadable copy, correction, deletion, restriction, or objection. Data export and account deletion tools are available in settings, and we may verify identity before acting. We may deny or limit a request where law permits.

The service is intended for people aged 18 or older. We do not knowingly seek data from a child under 18. If you believe a minor created an account, contact us so we can review and take appropriate action, including deletion when required.

We may update this policy as the product or law changes. We will post the revised version and effective date, and provide additional notice where a change is material or law requires it.

For privacy requests or questions, email support@efham.ai. This address is used pending owner verification of a dedicated privacy contact.