The challenge

Key Information

HMGCC Co-Creation will be hosting a two-stage competition process.

Phase 1. The objective is to rapidly assess one-page proposals, no feedback will be given at this stage. Successful applicants will be invited to phase 2. For further information please see How to apply – phase 1.

Phase 2. Successful phase 1 applicants will be requested to submit a proposal directly to [email protected]. For further information please see How to apply – phase 2.

Context of the challenge

National security engineering organisations regularly produce bespoke software for specialised deployments. This includes deploying software onto modified hardware, including mobile phones.

These devices are often used in critical missions, where failure is not an option. Testing on deployed devices must be thorough and robust. Industry will often do this using phone testing farms. We know employing AI agents could make this task easier.

This challenge is focussed on deploying AI agents onto mobile phones to replicate normal phone usage at scale.

The gap

Agent AI can already imitate human behaviour in many domains and work with a degree of autonomy for some tasks. Many sectors are looking to integrate these systems into workflows.

But few have been deployed on mobile devices. We need an AI agent that can mimic human behaviour patterns. It must be able to:

• Connect to a network
• Have some pre-trained personas to mimic a variety of different natural behaviours, including accessing web pages and sending messages.
• Replicate those habits at human-like speed, not machine-speed.

Example use case

Alayna is a team leader in a software development team creating bespoke software for mobile phones.

As part of testing and verification, she needs to build confidence that a phone still operates without bugs in a variety of different operating conditions, for example with varying user experiences and environmental conditions.

She uses a test environment in which different conditions can be simulated including GNSS positioning, network and WIFI access points, and environmental conditions such as temperature.

She also needs multiple phones to be used in different ways to simulate different use cases. Previously, this may have been done by test engineers virtually controlling these phones with pre-defined use cases – but what about when large numbers of phones have to be tested at once?

Instead, Alayna installs MIMIC (Mobile Intelligence Managing Individual Conduct) – an AI agent on each phone. Different personas can be selected which simulate real-world use cases. Each phone is used exactly as it would be by a person, from accessing website, using map services, to sending messages. With MIMIC installed, realistic phone usage is replicated at a speed never achieved before.

Project scope

Develop an AI Agent tool that can operate autonomously on a mobile device (or virtual surrogate) to mimic a person.

The outcome should be a demonstrator after this 12-week project, to a minimum Technology Readiness Level (TRL) 5 – technology basic validation in a relevant environment. The software should be demonstrated to the sponsor with a detailed report.

Essential requirements:

• The AI Agent should be demonstrated operating in an emulated environment.
• Must provide source code and a detailed technical report.
• Installation and operation must require no technical expertise.
• The agent runs unsupervised in training; a human can intervene during operational use.
• The project must demonstrate appropriate guardrails, including explainability of AI agent autonomy and robustness of cybersecurity.
• Must be able to send SMS, use messaging apps and place voice calls.
• Solution must make it possible for variety of personas to be selected.
• Able to route traffic through configurable VPN exit nodes.

Desirable requirements:

• Run on a physical mobile phone (not just an emulation).
• Target the smallest, most powerful efficient model available.

Constraints:

• Testing will take place in an on-premises environment.
• No training data will be provided.

Not required:

• Natural language phone conversations.
• Horizon scanning.
• Low TRL research

Eligibility

This challenge is open to sole innovators, industry, academic and research organisations of all types and sizes. There is no requirement for security clearances.

Solution providers or direct collaboration from countries listed by the UK government under trade sanctions and/or arms embargoes, are not eligible for HMGCC Co-Creation challenges.

Invitation to present

Successful phase 2 applicants will be invited to a pitch day, giving them a chance to meet the HMGCC Co-Creation team and pitch the proposal during a 20-minute presentation, followed by questions.

After the pitch day, a final funding decision will be made. For unsuccessful applicants, feedback will be given in a timely manner.

Clarifying questions

Clarifying questions or general requests for assistance can be submitted directly to [email protected] and  [email protected] before the deadline with the challenge title as the subject. These clarifying questions may be technical, procedural, or commercial in subject, or anything else where assistance is required. Please note that answered questions will be published to facilitate a fair and open competition.

How to apply

Please submit your application on the HMGCC Co-Creation website. Any queries please email [email protected] and [email protected].

Please refer to the information document linked below for further scoring criteria for phases 1 and 2.

All information you provide to us as part of your application will be handled in confidence.