MEMEX AI analyzes the notifications received on a vulnerable person’s smartphone to detect signs of risk - such as manipulation, scams, distress, or changes in behavior. If something concerning is detected, it sends a discreet alert to a carer or loved one via WhatsApp.

No. MEMEX AI does not read entire conversations. It processes notification summaries, which include inbound content only, using privacy-preserving AI. Personal content is never stored, shared, or viewed by humans.

Absolutely not. MEMEX AI is designed to respect the dignity and autonomy of the person being protected. It doesn’t track location, record calls, or allow someone to spy. It simply monitors for signs of potential harm. It does not enable anyone (including the guardian) to read the dependants messaging, it can only give you a heads up to check in about a potential threat.

All alerts are sent directly to the carer or guardian via WhatsApp. This keeps things simple and familiar - no new apps to download or check.

MEMEX AI can detect:

• Unusual requests for money or changes to legal documents

• Potential online scams or phishing attempts

• Signs of distress or isolation based on tone or frequency of communication

• Sudden behavioral shifts that could indicate injury or confusion

Each alert includes a short explanation and suggestion next step, without revealing private messages.

Yes. Transparency is important. Carers are encouraged to have an open conversation before setup. MEMEX AI is designed to support - not deceive - and does not run in secret. It is important for them to understand MEMEX AI;

• Maintains their independence and autonomy

• Is invisible and doesn’t require interaction

• Respects privacy - no names or personal data shared. No one can read what they do.

Setup requires a one-time installation on their device by a trusted person. After that, they don’t need to interact with the app - it runs quietly in the background without interrupting them.

Yes. All data is pseudonymized and processed securely. No personal names, message content, or sensitive details are shared. We take privacy and data protection seriously.

• £4.99/month for 1 dependant

• + £1/month for each additional dependant

• Custom pricing for enterprise partners like care homes or charities

You can pause or cancel anytime. If you stop using MEMEX AI, no further alerts will be sent and no data will be retained.

No system can promise to catch everything - and neither can we.

Digital communication and manipulation tactics are evolving rapidly, as are the tools used by scammers and bad actors. While MEMEX AI is trained to detect many common signs of emotional manipulation, phishing, and behavioral changes, we cannot guarantee that every threat will be flagged.

MEMEX AI is not meant to replace human care, but to augment your awareness. Think of it as an extra pair of ears—one that never sleeps—but not a perfect shield. We are constantly updating our models to adapt to new threats, and your feedback helps us improve every day.

Currently, alerts go to one WhatsApp number by default, but we are working on supporting multi-carer access, so siblings or care teams can share responsibility. You’ll be able to add others soon.

MEMEX AI may occasionally generate false positives—alerts triggered by messages that aren’t actually harmful. We’d rather err on the side of caution, but we’re constantly refining the system to reduce noise.

MEMEX AI is not surveillance software. It doesn’t track movement, block content, or record activity.

We don’t believe in control - we believe in care. Everything is designed to preserve the dignity, autonomy, and privacy of the vulnerable person.

Notification text is processed using AI models through encrypted channels. But critically:

• Data is anonymized and pseudonymized

• It is not linked to any account or personal identity

• It is never stored after analysis

Even OpenAI’s system sees a fragment of a message .- but never who it’s from, or what device it came from.