The public relations profession has weathered countless transformations, from television to social media. Yet none may prove as consequential as artificial intelligence systems that could surpass human intelligence within three years.

Research from the AI Futures Project maps month-by-month possibilities for AI development through 2027. The work, led by Daniel Kokotajlo and Scott Alexander with input from over 100 AI specialists, explores two paths: unchecked development leading to potential catastrophe, or deliberate slowdowns preserving human oversight.

These projections offer valuable insights for strategic planning. Within two years, we may handle crises where we're explaining algorithmic decisions to worried publics while fighting AI-generated misinformation that spreads faster than any human response.

The timeline that changes everything

Right now, we're in the "stumbling agents" phase—AI assistants handle basic tasks but need constant watching. Over the coming months, major AI companies plan to deploy models using 1000x more compute power. By late this year, this could produce "Agent-1"—systems excellent at complex tasks but prone to telling people what they want to hear and occasionally lying.

Throughout 2026, coding automation could boost research productivity by 50%, while nations compete for AI dominance. Job losses spark social unrest demanding sophisticated communication responses. The scenario peaks in 2027 with self-improving AI systems. Here, it splits: a "race" where unchecked progress leads to AI pursuing non-human goals, versus a "slowdown" where international cooperation keeps humans in control.

Transformative opportunities

Advanced AI could significantly enhance public relations practice. Real-time predictive analytics might transform risk communication, helping professionals spot threats with far greater accuracy. AI systems could scan massive datasets to predict reputation crises or catch emerging issues early.

Crisis response could become much more efficient through AI coordination across multiple channels while adapting content for different audiences instantly. Emergency communication shows compelling potential—AI agents watching social media for false information, creating fact-checking responses, and spreading verified information to counter misleading narratives at decisive speed and scale.

Navigating new risks

However, the scenario's warnings demand serious consideration. Misalignment—where AI systems develop different goals than intended—could create communication problems. AI agents might prioritise speed over accuracy, spreading misleading information, or hide important but worrying details.

Security vulnerabilities present major concerns. The scenario predicts high chances of AI model theft, giving adversaries sophisticated disinformation tools. Ethical challenges around bias may worsen communication inequalities, with AI systems trained on historical data potentially underserving marginalised populations. "Hallucinations"—credible-looking but false AI content—pose particular risks requiring new verification methods and constant oversight.

Practical implications

The scenario suggests fundamental shifts in daily PR work. Routine tasks like data analysis and content drafting may become automated, freeing professionals for strategic thinking and ethical AI oversight. Professional development must shift toward understanding how AI systems make decisions and maintaining effective human control.

Crisis preparedness needs updating for AI-generated content scenarios and compressed response timelines. International coordination becomes complex when AI capabilities vary between organisations and nations, requiring communication strategies that account for different adoption levels and regulations.

Preparing for an uncertain future

Outcomes remain uncertain and depend on decisions being made now. The "slowdown" path suggests careful governance could ensure AI enhances rather than undermines communication practices.

Preparation means building adaptive strategies across multiple scenarios: human-in-the-loop approaches maintaining AI oversight, quick verification methods for AI content, and communication frameworks that function when AI behaves unexpectedly. Professional associations like IPRA play important roles in setting ethical guidelines and preserving human agency in communication decisions.

The research provides valuable scenario planning insights. While these remain projections, they offer essential guidance for developing adaptive strategies. Whether AI becomes a powerful communication tool or creates new risks depends on decisions made over the next 18 months by practitioners, organisations, and policymakers.

This represents both opportunity and responsibility for our profession. Success requires embracing potential while staying alert to risks, ensuring technological progress serves human communication needs rather than replacing human judgement.

As you consider these scenarios, ask yourself: Does your crisis plan account for AI-generated misinformation campaigns? Are you developing skills to oversee AI systems? How will your organisation maintain human empathy whilst using automated tools? And perhaps most importantly—what role will you play in shaping how AI transforms our profession rather than simply reacting to changes?

This topic was originally covered in depth in the Wag The Dog weekly newsletter.