For Immediate Release is a two-person conversation between Neville Hobson and Shel Holtz about the news that makes a communicator's job harder. It calls itself the podcast for communicators, and the framing earns the label. A story about gig workers turns into a question about what employees should be told when their data trains AI models. A Hootsuite trends report turns into a question about whether brands can keep reacting to cultural moments within hours without wearing out the people doing the posting. That move from headline to consequence is the show's signature. It comes in two formats. Short midweek episodes each take one topic, while a monthly long-form edition covers several. In 'FIR #506: Battle of the Bots!' the list includes Anthropic's chief counsel forecasting the end of billable hours, Gartner predicting that AI-surfaced earned media will double PR budgets, and J.P. Morgan monitoring junior staff hours, which the bank says is to stop them working too much. The hosts go through the comments under the FT's post on that story, where one reader calls it 'a classic case of measuring what's easy instead of what matters', and then ask what is really being measured: time, activity, presence or value. AI keeps coming back, but the hosts' position changes with the case. 'FIR #507: Should Nobody Really Ever Write with AI?' argues that a chemist or engineer can write better reports with AI than without it. It also looks at the book Winning with AI, whose authors used Claude, ChatGPT, Perplexity and Gemini for research and structure while keeping the ideas, voice and judgment human. 'FIR #504: When Companies Blame Layoffs on AI -- and Leave Communicators Holding the Bag' is much sharper. It weighs Block's announced 40% workforce reduction and Oracle's cuts tied to spending on AI infrastructure against Anthropic's labor market research. Copyright gets the same practical treatment, with the hosts suggesting communicators start registering the content they produce. There are lighter moments too, like the idea that communicators need a worry list of their own. The advice stays concrete: disclose how employee data is used, write AI governance policy, name subject-matter experts staff can call, and plan in advance which stakeholders will need to hear what. Every episode signs off with 'that'll be a 30'. For a first listen, the AI washing episode shows the method most clearly: a headline, the data behind it, and the communicator stuck in the middle.