Source [De]Coded — The platforms changed. Did your strategy?

Morning all,

London is having a moment. Our UK team just picked up two shortlist spots at the PRCA UK Awards, one in the specialist categories and one in the size categories. They also hosted a panel this week on the UK’s alternative media landscape. Two nods and a full room in the same week? Somebody please tell London to save some trophies for the rest of us.

Speaking of the future, nearly every PR pro is now using AI. That means using it isn’t the flex anymore. This week we’re digging into why the real question isn’t whether AI replaces your juniors but what it lets them do on day one. We also look at how Discord is opening its ad business to smaller brands, why state courts are writing the social media rules Congress won’t, and how AI shopping agents are quietly picking winners before most brands even know the criteria. The middlemen are changing, and fewer of them are human.

Let’s get into it.

Becky and Greg

Stop Asking If AI Will Replace Your Juniors. Ask What It Lets Them Do on Day One.

With nearly every PR pro now using generative AI, the agencies pulling ahead are the ones building tools around their own standards and training people to use them well. Read the full piece on The Source.

TL;DR:
According to Cision’s Inside PR 2026 report, 91% of PR professionals use generative AI in their workflow, mostly for idea generation (73%) and writing (68%). The industry conversation still fixates on cutting headcount, but the better question is what happens when a first-year account coordinator has the same research, writing, and audit tools as a VP. SourceCode’s answer: build the stack on your own voice, research, and audit standards, design it around the EU AI Act and PR Council guidelines, and invest heavily in training.

Takeaway:
Adoption is no longer the differentiator. When everyone has the same off-the-shelf models, generic tools produce generic work, and the advantage moves to whoever encodes their own judgment into the system. That changes the talent equation. The value of AI isn’t making senior people a little faster. It’s raising the starting point for everyone else, so junior output begins where senior thinking would. But that only holds if governance and training come with the rollout. Agencies and in-house teams that hand people powerful tools without guardrails are taking on compliance and reputational risk, especially with EU transparency obligations now in force. The teams that win will treat AI as an operating standard, not a productivity hack.

Consider:

• For marketers: Are your AI tools trained on your brand voice, research methods, and quality bar, or are you measuring speed gains on output that still needs senior rework before it ships?
• For communications leaders: Is your leadership team framing AI internally as a headcount story or a capability story, and what is that framing doing to how your newest people learn, grow, and use these tools responsibly?

Discord Opens Its Ad Business to Smaller Brands With New Self-Serve Tools

Discord is making it easier for a wider range of advertisers to buy on the platform, with a self-serve manager and expanded video formats.

https://arstechnica.com/gadgets/2026/09/discord-gives-advertisers-more-tools-to-target-its-users

TL;DR:

Discord has launched Ads Manager, a self-serve campaign management platform, and expanded its Video Quests product to reach a broader advertiser base. Together, these moves open Discord’s ad inventory to smaller brands that may not have had access or resources to buy through managed channels.

Takeaway:

Self-serve tools are how platforms signal that they are serious about ad revenue at scale. Discord has a genuinely distinctive audience, particularly in gaming, tech, and younger demographics, and lowering the barrier to entry gives more brands a reason to test it. For marketers, that is an opportunity to experiment before the platform gets crowded and expensive. For communications leaders, Discord’s ad expansion is also a reminder that community-oriented platforms are increasingly becoming ad-supported ones, which changes how brand presence there is perceived by users.

Consider:

• For marketers, is Discord on your experimental media plan for the next cycle, and if not, what is the reason given the lower barrier to entry?
• For communications leaders, how do you balance a brand’s organic community presence on Discord with the introduction of paid advertising on the same platform?

States Are Setting Advertising Rules That Congress Has Not

Settlements and rulings from state-level legal actions against major platforms are quietly reshaping how social media advertising works.

https://martech.org/states-are-starting-to-rewrite-the-rules-of-social-media

TL;DR:

State-level legal actions against Meta, TikTok, and YouTube are producing settlements and rulings that are setting new operational and advertising rules for those platforms. In the absence of federal legislation, state courts and attorneys general are becoming the effective regulators of social media.

Takeaway:

When regulation comes through litigation rather than legislation, it tends to be uneven, unpredictable, and cumulative. Marketers who rely heavily on targeting capabilities tied to user data on these platforms should be paying attention, because the rules governing what is permissible may vary by state and shift further as more cases resolve. For communications leaders, the reputational dimension matters too: being associated with a platform that is actively defending itself against state attorneys general carries brand risk that is separate from any direct regulatory exposure your company faces. The strategic posture here is not panic but it is not complacency either.

Consider:

• For marketers, have you mapped which of your targeting practices on major platforms are most exposed to the kinds of restrictions emerging from these state-level actions?
• For communications leaders, does your brand have a clear internal position on continued investment in platforms that are under active legal scrutiny from multiple states?

AI Agents Are Choosing Products for Shoppers, and Most Brands Are Not Ready

New data shows AI agents are stepping between brands and buyers, and the criteria they use to make selections are largely invisible to marketers.

https://digiday.com/marketing/in-graphic-detail-where-brands-stand-when-ai-agents-do-the-shopping

TL;DR:

Data examined by Digiday illustrates how AI agents are increasingly acting as intermediaries in the shopping process, selecting products on behalf of users. Most marketers do not yet understand the criteria these agents apply, nor do they have strategies in place to influence them.

Takeaway:

The premise of most marketing is that you can reach the person making the decision. AI agents complicate that premise by inserting a layer of algorithmic judgment that responds to inputs most brands have not optimized for. This is not a distant scenario: the data suggests it is already affecting where products appear and whether they get chosen at all. For marketers, the urgent question is what signals, structured data, content formats, or platform relationships actually influence how agents evaluate and recommend products. For communications leaders, this is also a brand visibility issue: if agents are filtering on criteria like reviews, return rates, or structured product information, then comms and marketing need to be working from the same playbook.

This is what some are calling Assistive Agent Optimization (AAO): the shift from being discovered and described by AI to being selected by it. For more on what that means for B2B brands, see our recent piece, AAO: WTF Is It, and What B2B Marketers Actually Need to Know.

Consider:

• For marketers, do you know what signals AI shopping agents are using to evaluate your products, and are you actively optimizing for any of them?
• For communications leaders, is your team coordinating with marketing and e-commerce on the structured content and reputation signals that AI agents appear to weight heavily?