Reflections from mumbrella commscon

At Mumbrella CommsCon, one message came through clearly: AI is no longer emerging, it is already transforming communications.

Across panels, keynote speeches and Q&As, the same challenge came up. How must communications adapt to remain effective in an environment where influence, trust, and visibility are increasingly shapes by algorithms?

For pharma, this is important. Reputation directly influences patient trust, treatment decisions and regulatory confidence, and AI is shaping how that reputation is formed, surfaced, and scrutinised.

Three shifts in communications

where trust is built is evolving

For years, communications followed a relatively predictable model. Word of mouth was the most influential driver of brand perception, grounded in lived experience and human validation. Public relations (PR) played a vital role in shaping that through earned media, which then flowed into broader public discourse. Social media extended the reach.

Today, audiences are turning to decentralised, interest-led “new media” ecosystems including podcasts, Substack newsletters, blogs, streaming platforms, and community-based platforms like Reddit to talk about their health, experiences and treatment options. These are not broadcast environments. They are participatory, self-regulating, and built on relevance and authenticity rather than reach.

Critically:

• Word of mouth hasn’t declined in importance, but it has been digitised, made searchable and reach larger audiences than ever before
• Influence is now shaped by collective interpretation – whether an online campaign succeeds is not determined by number of engagements, but instead is based on the collective view from the community
• Trust is built within communities, not by brands pushing their content out

Reddit is a good example. Its anonymity strips away performative behaviour, creating space for more candid, experience-led conversations. For pharma, that’s both valuable and uncomfortable. These conversations are happening, with or without you.

At the same time, the media landscape is becoming increasingly fragmented and decentralised. Nearly half of journalists are now publishing, or considering publishing, on platforms like Substack, prioritising direct relationships with audiences and greater editorial control. As a result, influence is no longer concentrated within traditional media outlets, but it is dispersed across a growing number of independent voices and platforms.

This shift has significant implications for visibility in the age of AI. With up to 89% of AI-generated citations drawing from earned media, and ‘new media’ platforms like Reddit and LinkedIn actively informing how algorithms are trained, influence is now shaped across a far broader digital ecosystem.

It is no longer enough to secure coverage in a single channel. To be effective, brands must coordinate presence across paid, earned, shared, and owned media (recognising that shared platforms increasingly involve ‘new media’ and community forums) to ensure they are consistently cited, referenced, and discussed.

What should brands do next to build trust?

1. Recognise that ‘new media’ and community platforms are no longer ‘fringe’ channels but influential parts of the communications ecosystem which are now shaping both patient perception and AI-generated narratives.
2. Consider establishing clear governance for listening, engaging (where appropriate), and responding in community spaces to help shape narrative early, mitigate misinformation, and build authentic credibility within patient communities.
3. Design campaigns that deliberately connect earned media, social, shared (community) and owned content, ensuring messages are reinforced across each touchpoint.

How reputation is tracked and managed is changing

Reputation has traditionally been reviewed through static snapshots, comparing moment in time metrics. That model is no longer fit for purpose.

AI has changed both the speed and the stakes. It’s now possible to track how your organisation, products and therapy areas are being discussed in real time across media, social platforms, community forums and AI-generated search. More importantly, those signals can be linked directly to commercial and risk outcomes.

At the same time, AI is reshaping how people access information. Whether through Google’s generative search or tools like ChatGPT, audience are instantly accessing a synthesis of everything that already exists in the public domain, including resurfaced narratives about brand credibility and controversy.

And critically, AI does not forget. Historical statements, outdated data, or inconsistencies can be resurfaced instantly. If inaccurate or incomplete information exists, it can, and will, be used to form answers. This introduces a new category of risk.

With one panellist stating that organic search is projected to decline by 70% by 2027, Generative Search is already a primary source of information for many Australians. “Moment-in-time” reputation tracking is therefore no longer enough. If your organisation is not actively monitoring, managing, and shaping how it appears across AI and community ecosystems, you are outsourcing your reputation to algorithms.

What should brands do next to manage reputation?

1. Establish always-on monitoring across AI search, traditional media, social and community platforms to understand how your organisation, products, and therapy areas are being represented in real time.
2. Audit your “AI footprint” to identify misinformation, outdated content, or gaps and respond with coordinated communications that correct and reinforce accurate narratives.
3. Connect reputation insights to commercial and risk functions so teams can respond quickly and confidently when narratives shift.

The faster you create media materials, the less journalists trust it

Medianet launched their annual Australian Media Landscape Report at CommsCon which offered a useful reminder. Whilst AI is increasing the volume of content, the bar for what is trusted is raising, and we must work to continuously meet those high standards.

From more than 800 Australian journalists surveyed:

• 54% now use AI (up from 37% in 2024), primarily for summarising content, transcription, and research, yet 93% are concerned about its impact on quality and honesty.
• 73% receive AI-generated pitches and 78% say they lose trust if they suspect a pitch is AI-written, citing poor grammar, incorrect data, and even wrong contact details as the main red flags.
• 66% see PR as critical to their success, while 86% rely on media releases as a primary source of information, more than direct contacts.

In a media landscape flooded with low-quality, AI-generated slop, well researched, written and referenced media pitches, which offer genuine ‘new news’, visuals/assets and top tier spokespeople come out on top. Speed has not, and will not, replace credibility and accuracy.

What should brands do next to maintain journalist relationships and trust?

1. Ensure all outward-facing content, especially media materials, are accurate, well-written, and genuinely valuable, with human oversight guiding AI-supported workflows.
2. Take the time to develop clear, evidence-based, and newsworthy media releases that can be reliably referenced, reused, and surfaced over time.
3. Remember efficiency gains are meaningless if they erode trust, so lead with credibility, not convenience.

The bottom line

The question is no longer whether AI will change communications. It already has.

If you’re unsure how your organisation is showing up across media and generative AI, reach out to our expert consultants.

Contact Cube