April 2026

Release Date: 24 April 2026

What's in This Release

This release marks a step change in how exercises are built and run: AI-powered cognitive modelling makes personas and factions more convincing, while the new Communications Manager lowers exercise control costs by reducing role players needed to run them.

Reduced exercise control costs

  • Communications Manager reduces the friction of role-playing multiple stakeholders, it introduces the option of AI-responses from those stakeholders either with a human-in-the-loop or with auto-response. This works across email, website articles and social media.

Improved realism

  • Personas and Factions now have “cognitive modelling” meaning that these stakeholders have goals, beliefs and attitudes that means the AI generates better content and can respond in a credible way to exercise events.


Improvements & New Features

Communications Manager

What's New

The Communications Manager is a new enhanced capability in Social Watch that gives role players and exercise controllers a single, unified workspace to manage persona impersonation across social media, email, and articles - consolidating what was previously spread across Social Watch and multiple channel views into one command centre.

Why it matters

  • Faster, simpler exercise control

  • Clear visibility of all persona activity in one place

  • Easier shift handovers with no loss of context

  • One person can manage multiple personas across multiple channels simultaneously

  • AI responds in character on behalf of personas, reducing the headcount needed to run exercises

Licence

Eagle

See full details and examples on Communications Managerarrow-up-right

Key features

Unified impersonation feed

Composer (single publishing interface)

AI-assisted responses

Autonomous Personas

MIND

  • One chronologic AI view per persona

  • Shows posts, messages, replies, and AI responses together

  • No need to switch channels or track activity manually

  • Send emails, posts, and articles in one place

  • Reply mode auto-fills recipients

  • Includes relationship insights: trust, disposition, power, interaction history

  • Describe the situation → AI generates in-character content

  • Edit before publishing

  • Uses full persona cognitive model for realism

  • Personas can respond automatically using AI

  • Maintains character using beliefs, style, and memory

  • All activity remains visible in the same feed

  • Persona memory summary

  • Current mindset

  • Key facts

  • Gives an audit trail of how the persona is evolving

exclamation

Important: For Autonomous Personas to generate accurate, in-character responses, two things must be in place before the exercise runs: the scenario description must be completed in the Exercise Editor (this populates the AI's exercise context), and each persona's cognitive model must be fully built out. Personas must also be assigned to positions, and those positions assigned to role players, for impersonation options to appear.

chevron-right ⚙️ How It Workshashtag
  1. Open the Social Watch channel in the Player View.

  2. Select the Impersonation option and a persona you have permission to impersonate. The unified feed loads, and as the exercise progresses, it will show all inbound messages, outbound content, replies, and AI-generated responses (if auto is enabled).

  3. To publish manually, click Composer and select the channel type - email, social, or article. Add your content and publish as the persona.

  4. To reply to a message, click Reply in Social Watch. The Composer opens in reply mode with the recipient pre-filled and the relationship attributes panel displayed.

  5. To use AI generation in the Composer, describe the event or context in the right panel. The AI generates a response; push it to the left panel, edit as needed, and publish.

  6. To enable Autonomous Personas, switch the impersonation column to auto mode. The AI will respond to incoming messages automatically, drawing on the persona's full cognitive model and exercise context.

  7. To review a persona's evolving state, click MIND to open the panel showing memory summary, mind state, and facts.

  8. To view or edit the exercise context used by the AI, open AI Settings and review the exercise context field. This is auto-populated from the scenario description in the editor.

circle-info

Use Case Example

A crisis communications exercise involves three active media personas simultaneously receiving and generating social content. Rather than assigning a dedicated role player to each, a single exercise controller opens the Communications Manager, monitors the unified feed across all three personas, and uses Autonomous Personas to handle routine incoming messages in character.

When a critical reply requires a human touch, the controller switches to manual mode, reviews the relationship attributes panel to understand the dynamic between the personas, and crafts a deliberate response - all from one screen, without missing a message.

Persona Cognitive Modelling

What's New

Personas in Conducttr now support a richer, structured set of cognitive and behavioural attributes, giving the AI the depth it needs to generate realistic, consistent, in-character content and responses across all channels. The Cognitive Model powers AI generation in two key places: in the Communications Manager, where it drives in-character responses and Autonomous Personas during live exercises; and in the Exercise Editor, where the in-editor AI draws on it to generate persona content while building the exercise.

Why it matters

  • The AI generates more realistic, consistent, and in-character content across all channels

  • Personas become functional data models, not just text descriptions

  • Speeds up persona cognitive model creation with AI auto-fill from the persona's Bio

Licence

Core - All licences

Expert - Eagle

Learn more about Persona Cognitive Modellingarrow-up-right

Key features

Core view

Expert view

AI auto-fill (Full Cognitive Model profile)

AI auto-fill (Communication Style)

Essential attributes for realistic AI output, available to all users: identity, background, beliefs, communication style, tone, delivery, and more. Includes tooltips on every field

Full attribute set for users who want deeper control, including advanced fields

One click fills all persona attributes from the persona's Bio in both Core and Expert mode.

Targeted auto-fill for the Communication Style section only, triggered from within that section

memo-pad

Changes in any view (Core or Expert) update the others automatically.

chevron-right⚙️ How It Workshashtag
  1. Open a persona in the Exercise Editor and navigate to Persona Details.

  2. Use Core view to fill in essential attributes: write the persona Bio and use the AI to generate the full cognitive model - identity, background, beliefs, communication style, tone, delivery, and more. Tooltips on each field explain its purpose and effect on AI generation. Review and adjust the generated values as needed.

  3. Switch to Expert view for the full attribute set. Eagle-licence users can access this area and copy the persona's complete schema.

  4. To auto-fill Communication Style only, use the AI button within the Communication Style section. Existing values will prompt a confirmation before being overwritten.

  5. Ensure personas used for impersonation or Autonomous Personas have their cognitive model fully completed - this directly affects the quality and accuracy of AI-generated responses.

circle-info

Use Case Example

A designer is building an exercise where personas will be impersonated using the Communications Manager and where the in-editor AI will also be used to generate content. For each persona, they write a short bio and click AI auto-fill to generate the full Cognitive Model, populating beliefs, communication style, tone, and more in one step. They then fine-tune individual fields where needed. With the cognitive model complete and the exercise setup configured in Configurations, both the in-editor AI and the Communications Manager can generate content that is accurate, consistent, and genuinely in character for each persona.

Faction Library

What's Improved

Factions are groups that personas belong to, such as a government agency, a media outlet, a militant group, or a public audience. They give personas a shared identity, communication style, and set of beliefs.

Factions have now it’s own dedicated library, with a structured table view, new search and filtering capabilities, and expanded editing options - making it faster and clearer to browse, create, and manage factions at the start of exercise design.

Why it matters

  • Faster, clearer workflows for creating and editing factions

  • Structured faction-first design process aligns with how realistic information environments are built

  • Cognitive and influence attributes for Audience-type factions improve scenario realism and analytical depth

  • Flexibility for both basic and advanced users through Core and Expert views

Licence

Core - All licences

Expert Mode - Eagle only

Key features

Faction Library table

Search & filter

Core & Expert tabs

Audience cognitive attributes

New organised and dedicated view showing name, description, empathy, and A3E (Audience, Actor, Adversary, Enemy)

Search by name or description, filter by A3E type

  • New interface for all standard faction attributes and communication style settings

  • Editable JSON tree for users who want direct control over the full faction schema

Dynamically shown only for Audience-type factions: Capability, Opportunity, Motivation, Mindset, Belief System, and Behaviour

chevron-right⚙️ How It Workshashtag
  1. Open the Faction Library in the Exercise Editor.

  2. Browse factions in the structured table. Use the search bar to filter by name or description, or use the A3E type filter to narrow by Audience, Actor, Adversary, or Enemy.

  3. Click Add Faction to create a new faction - the editing panel opens directly.

  4. In the Core tab, complete the standard faction attributes including name, description, empathy level, A3E classification, and communication style.

  5. If the faction is classified as Audience, the cognitive and influence attributes section appears automatically. Complete Capability, Opportunity, Motivation, Mindset, Belief System, and Behaviour as required.

  6. For advanced users, switch to the Expert tab to edit the faction's full JSON schema directly

circle-info

Use Case Example

A designer building a counter-disinformation exercise needs to model a susceptible public audience as a faction before creating any personas. In the updated Faction Library, they classify it as Audience type, set the empathy slider, and fill in the cognitive attributes - Motivation, Belief System, and Mindset. When building personas that belong to this faction, if a persona is missing attributes such as beliefs or communication style, the AI will fall back to the faction-level attributes to generate content, ensuring the output remains coherent and in character even for personas that haven't been fully built out yet.

Bug fixes

Bugs have been resolved across the platform to improve stability and overall user experience.

Related content


Release Date: 11 April 2026

What's in This Release

This release improves opportunities for clients to increase their differentiation and confidence while building and running exercises.

Increased differentiation

  • Inbound API enables client applications to connect to live exercises.

  • BYOP (Build Your Own Prompts) allows clients to create custom AI prompts for After Action Review.

Increased confidence

  • Exercise Verifier checks your exercise to find possible configuration mishaps before you publish.

Improvements & New Features

Inbound API

What's New

Conducttr now supports Inbound APIs, allowing external systems, custom tools, and automation workflows to connect directly to live exercises. Developers and technically capable teams can use the API to send messages, publish content, control exercise timing, and retrieve data - without manual intervention in the Facilitator Dashboard, Player View, or Editor.

Why it matters

  • Enables integration with external platforms, data feeds, and custom tooling

  • Automates inject delivery and exercise control without manual facilitator action

  • Opens up new exercise design possibilities: scoring systems, live data feeds, and more

  • Full documentation, endpoint references, use cases, and code samples available in the Conducttr Help Docs

Key features

  • Send messages to teams, positions, and roles

  • Manage Smartwords

  • Retrieve persona and team data

  • Publish articles

  • Control Pattern of Life (POL) - start, stop, and query current injects

Licence: Eagle

chevron-right⚙️ How It Workshashtag
  1. Review the API Core Conceptsarrow-up-right to understand authentication, base URLs, and request structure.

  2. Browse the Endpoint Referencearrow-up-right to identify the endpoints relevant to your use case.

  3. Use the Possible Use Casesarrow-up-right guide for inspiration on integration patterns.

  4. Follow the code samplesarrow-up-right to get started quickly with common tasks such as publishing articles to teams or running a quiz and dice demo.

  5. Test your integration against a live exercise environment and validate responses using the endpoint reference documentation.

circle-info

Use Case Example

A government training team wants to integrate a real-time threat intelligence feed into their cyber exercise. Using the Inbound API, they build a lightweight script that monitors the external feed and automatically publishes relevant articles into the exercise as Smartwords are triggered - injecting live-feeling intelligence updates into the player view without any manual facilitator action.

Exercise Verifier

What's New

The Exercise Verifier is a new checking tool built into the Exercise Editor that surfaces configuration issues before publishing - giving designers the opportunity to identify and fix problems while staying in the editor, rather than discovering them during a live exercise.

Why it matters

  • Catch configuration mishaps before they become live exercise problems

  • Fix and re-check iteratively without leaving the editor

  • Consciously accept issues you choose not to action

  • A summary also appears at the publishing step as a final review prompt

Full details on all checks are available in the Verifier documentationarrow-up-right.

Licence: All licences

chevron-right⚙️ How It Workshashtag
  • In the Exercise Editor, click the Verifier icon next to "Designing for" — this is visible throughout the editor at all times.

  • The Verifier panel opens. Click Run check to scan the exercise for configuration issues.

  • Review the list of issues. Each issue includes a description and guidance on where to go to fix it.

  • Navigate to the relevant area of the editor, make the fix, and return to the Verifier panel.

  • Click Run check again to confirm the issue is resolved.

  • For issues you choose not to action, mark them as accepted.

  • At publish time, a summary of any outstanding Verifier issues is shown in the publishing step as a final prompt to review.

circle-info

Use Case Example

A designer has built a complex multi-channel exercise and is preparing to publish. Before clicking publish, they open the Verifier and run a check. It surfaces two issues: a website tab not configured to open automatically, and a channel set to open on first inject with no content. The designer fixes all directly in the editor, re-runs the Verifier to confirm all clear, and publishes with confidence, avoiding issues that would previously only have appeared mid-exercise.

BYOP - Build Your Own Prompts

What's New

Exercise designers can now create custom AI prompts for use in After Action Review (AAR), enabling teams to apply their own analytical frameworks to player responses rather than relying solely on Conducttr's built-in prompts.

Why it matters

  • Apply your organisation's own assessment criteria to player responses using AI

  • Custom prompts appear alongside Conducttr's built-in AAR prompts

  • Prompts can be tested before the exercise runs

  • Prompts carry over when a scenario is uploaded, keeping setup consistent across runs

See full details and examples on BYOP documentationarrow-up-right.

Licence: Eagle

Key features

Feature
Detail

Up to 3 custom prompts per exercise

Managed under Configurations > Exercise > BYOP AAR

Prompt structure

Each prompt includes a name, description, and optional Content toggle - when enabled, the prompt appears in AAR only when analysing messages with contextual evidence (such as replies)

Test before saving

Enter a sample message to preview the AI's output; test content is not stored

AAR integration

Custom prompts appear in the AAR dropdown alongside Conducttr's default prompts

chevron-right⚙️ How It Workshashtag
  • In the Exercise Editor, go to Configurations > Exercise > BYOP AAR and open the Prompt Studio area.

  • Click Create prompt to add a new prompt.

  • Enter a Prompt Name and Prompt description.

  • Optionally enable the Content toggle if this prompt should analyse messages in context (e.g., replies with prior conversation). The Content field will then appear during testing.

  • Test the prompt by entering a sample message and clicking Test prompt. Review the AI output. Note: the test content is not saved.

  • Click Save to save the prompt settings. Toggle the prompt on or off to control which prompts appear in the AAR dropdown.

  • During the exercise, observers access custom prompts in the AAR dropdown when analysing a player response, alongside Conducttr's built-in options.

circle-info

Use Case Example

A government training team uses a proprietary strategic communication assessment framework. Using BYOP, their exercise designer creates three custom prompts that mirror the framework's criteria - evaluating audience awareness, message framing, and channel appropriateness. During After Action Review, observers select these prompts to run AI assessments against player responses, generating feedback that maps directly to the organisation's own evaluation standards - consistent, fast, and repeatable across every run of the exercise.

Further Improvements

Clear Teams List

Designers can now wipe the entire teams list in one go directly from the Exercise Editor, including all teams and participants. Previously, rows could only be cleared team by team, which was time-consuming for exercises with large numbers of participants or teams.

MEL Timer Persistence on Refresh

The current inject timer in the Facilitator's MEL view now persists through browser refreshes. Previously, if a facilitator lost connection and refreshed the page, the timer would reset to the beginning. The timer is now stored locally and restored automatically on reload.

Instamedia - Image Now Required

Adding an image is now mandatory when creating an Instamedia inject, both in the Exercise Editor and in the Communicator. Previously it was possible to publish an Instamedia post without an image, which caused the layout to appear broken and distorted for players.

Bug Fixes

Bugs have been resolved across the platform to improve stability and overall user experience.

Last updated

Was this helpful?