Did you see it coming?

The fact that OpenAI has been recruiting for a new head of preparedness since the turn of the year has caught the attention of many in the media business, not least because of the remuneration package of $550,000 salary plus generous share options.

The job itself is not actually new. It was first created in 2023 when OpenAI first formed its preparedness team. Yet the role fell vacant in mid-2024 after leadership changes within the company.

Now the job has come back, but in a slightly new form. It now covers:

• Predicting dangerous capabilities of their own new AI products.
• Building systems to measure them.
• Designing safeguards and “guardrails”, while anticipating “jailbreaks” when the AI tools become too smart, often unexpectedly.
• Deciding when a model is safe enough to release.
• Protecting the whole business from external cybercrime players.
• Coordinating activity across the organisation, including engineering, policy and leadership.

The current hiring process reflects a renewed emphasis on safety, ethics and mental health, amid lawsuits, public scrutiny, and rapidly advancing model capabilities. It is what OpenAI itself describes as “frontier-model catastrophic-risk preparedness”. Sam Altman wants a head of preparedness not for PR, hype or even growth, but for survival. Someone paid to imagine the worst before it happens.

So, three questions:

1. Do we as publishers (or whatever we call ourselves now) need to have such a role?
2. What are our industry-specific threats that we need to prepare for?
3. Who in the organisation should do it?

Do we need to be prepared?

Yes, of course. And most of the larger companies used to do it in the past much better than they do now. Currently, there are loads of brilliant technicians who are trimming costs, renegotiating print contracts, consolidating vendors, squeezing budgets, etc. Yet genuine “Doomsday Thinkers” (with spreadsheets, scenarios and a PowerPoint slide deck) are few and far between.

What are we preparing for?

A publishing-sector version of preparedness would include:

• Understanding how users discover and consume content, especially in a changing, Google-zero world;
• Managing and protecting all the distribution pipelines for both physical and digital content as the platforms’ strategies constantly change;
• Protecting editorial trust and the direct relationship with end users;
• Safeguarding IP and licensing rights from scraping and training usage;
• Anticipating the impact of individual content creators;
• Strengthening synthetic-media detection;
• Anticipating AI-driven misinformation waves;
• Predicting the next generation of AI capability;
• Building crisis-response playbooks;
• Guiding responsible internal use of AI tools;
• Supporting commercial negotiations with AI vendors.

Whose job is it?

“It’s everyone’s job,” is a common response. Yet that is often a cop-out. Currently, most publishers have distributed responsibility across editors-in-chief (content quality); chief technology officers (infrastructure & security); chief financial officers; chief revenue officers. Or teams: legal & compliance; audience & SEO. A few have chief security officers, focusing solely on cybersecurity.

The Domesday Thinker must have: authority, granted from the top, with access to the CEO and board – that authority must include the power to convene cross-functional teams; the personal skills to challenge assumptions without being labelled “negative” and to go beyond buzz-phrases and marketing-speak; a dashboard or checklist of dependencies and vulnerabilities.

Or is it your job?

For a small, niche publisher, a highly paid Domesday Thinker is simply unaffordable. The alternative is to have the discipline (either on your own, with an internal team or an external group of partners) to look weekly, monthly and quarterly at changes in the external ecosystem, clearly defined dependencies and vulnerabilities across revenue, distribution and platforms. And to build what-if scenarios that lead to coherent response strategies.

Yet, preparedness is a state of mind and a discipline before it becomes a job or a title. It is not negative; it is essential for survival. And the positive flipside is that it will inevitably surface opportunities and beneficial pivots.

Most publishers have traditionally worked to very short time horizons: the next issue, the next ABC figure, the next budget cycle, the next financial announcement, etc. Ironically, the faster that change comes at us, the less we prepare. The result is that we simply will not “see it coming”, whatever “it” is.

So, if the accurate forecasting of the future is becoming almost impossible in a Trump world of surprises and curveballs, then being prepared for anything to happen is the next best thing, as any scout will tell you. Baden-Powell himself said back in 1908 that the phrase “be prepared” meant that a scout “is always in a state of readiness in mind and body to do their duty.” When asked “prepared for what?”, he famously replied, “Why, for any old thing!”

So, let’s join OpenAI and the scouts with “frontier-model catastrophic-risk preparedness”.

This article was first published in InPublishing magazine. If you would like to be added to the free mailing list to receive the magazine, please register here.