AI is undoubtedly transformative, but defining what this really means requires navigating a sea of over-hyped reports commissioned to help companies selling AI products and services.
The Thomson Reuters Future of Professionals Report 2025 stands out by calling out the hype and attempting to provide a report focused on action rather than hype. The report looks at the current landscape of AI adoption in professional services. It emphasises the importance of a strategic approach for organisations seeking to maximise return on investment (ROI) from AI technologies. The report critiques the hyperbolic sales messages surrounding AI, citing the need for accurate data and cautioning against inflated expectations of AI.
The key takeaway is it is critical to invest in AI, but also to have realistic expectations and always ensure human in the loop, highlighting the necessity for human oversight in AI outputs.
Already, professionals report increased efficiency, productivity, and cost savings as the most significant benefits, making AI a crucial tool for organizations navigating a rapidly evolving set of business challenges.
Thomson Reuters claims its own survey of professionals in the legal, compliance, and financial sectors found 53% reporting hard returns—largely in time savings and a reduction of errors.
The key themes to emerge from the Future of Professionals Report are AI’s transformative potential, data and ROI accountability, accuracy, disconnect in goals, work-life balance improvements, potential risks and future outlook.
It also identifies a four-layer governance model for AI success.
AI’s transformative potential
Generative AI promises to significantly impact professions like legal, compliance, and accounting, with professionals reporting increased efficiency and productivity. It is logical to extrapolate the same benefits to public relations, communications and corporate affairs.
This needs to be caveated, as automating codified professions governed by rules such as law and accountancy is easier than public relations, which doesn’t have the same agreed body of knowledge to draw on. This means more care and effort is needed to set up policies and create custom agents to focus on pre-qualified quality knowledge sources, which can vary dramatically from company to company.
AI ROI accountability
The Thomson Reuters report criticises many AI reports for lacking trustworthy data on ROI. The Future of Professionals report claims 53% of professionals in surveyed sectors experience tangible returns, especially in time savings and error reduction. It is particularly critical of over-hyped claims shared on social media.
Accuracy
For those critical of AI one of the most frequently cited problems is AI hallucinates, so its outputs might not be correct. I always remind people that AI is an assistant and just as you’d always check the work of a junior colleague, you always need to check the work of AI.
The report provides an explanation of why people often fail to do this. A significant majority (91%) believes that AI outputs should meet higher accuracy standards than human outputs, with 41% insisting on 100% accuracy before AI outputs can be used without review.
Disconnect in goals
There is often a mismatch between organizational AI strategies and individual goals, leading to inefficient adoption practices. We (Purposeful Relations) believe this is exacerbated by the slow pace of adoption by businesses. Individuals often experiment with AI tools personally and even use unauthorised tools for work. How they use it in these cases can be different to where the biggest benefits come if AI is adopted across a whole company or team.
Two thirds (65%) of respondents who said they have their own personal goals for AI adoption also said they aren’t aware of their organization having an AI strategy. This means they’re being asked to adopt new technology without critical guidance as to what they are trying to achieve for their organization by doing so. This disconnect makes it highly likely that such individual AI adoption will be inconsistent and inefficient.
Conversely, more than one-third (38%) with an AI strategy also reported that they don’t have any personal goals for AI adoption. This, too, increases the risk that the organization’s AI strategy cannot be effectively implemented if there is a lack of accountability and communication at all levels.
Work-life balance improvement
The report shows that AI can improve work-life balance by automating burdensome tasks and facilitating accelerated skills development.
And all of this AI use can contribute to an organization’s ability to attract and retain high-quality professionals.
Potential risks
While AI aids in time and cost savings, there are risks associated with industries that rely heavily on unverified web-sourced data, leading to potential errors and inefficiencies. There is so much dubious web content (and even books) about PR and communications that it is an even bigger threat than in professions with clearly defined rules.
AI introduces or exacerbates numerous reputational risks. Disinformation and misinformation were already a threat, but AI increases the risk exponentially as AI makes it easier and faster to create incredibly realistic fake news and social content, including deep fake videos and audio of senior executives.
It’s also critical to introduce governance to reduce risk and to help ensure social licence for AI so internal and external stakeholders understand and accept how it is being used fairly and ethically.
This is another reason why a strategic, planned approach is both more effective and safer. Creating a suite of AI uses that are grounded in valid, authorised data and knowledge both improves results and reduces risks.
Future outlook
There are numerous reports and case studies that show dramatic differences in time savings from AI adoption. One of the worst was an Australian government pilot which found Microsoft 365 Copilot saved just 14 minutes a week. The report also acknowledged the pilot still broke even and would have been more successful if there had been a plan and training. A similar pilot by the UK government – with a strategic plan and training – found Copilot saved an average of 26 minutes per day.
Most studies identify productivity savings of between two to four hours a week. The Thomson Reuters report is even more optimistic and says in 2025 legal professionals anticipate freeing up nearly 240 hours per year due to AI adoption, which is up from 200 hours in 2024.
Four-layer governance model for AI success
The report identifies that successful AI adoption requires a four-layer AI governance model comprising strategy, leadership, operational, and individual layers.
- Strategy: Organisations with clear, visible AI strategies are almost four times more likely to be experiencing immediate benefits than those without a clear, visible strategy for AI adoption.
- Leadership: Organisations with leaders who lead by example are seeing more benefit than organisations that don’t have visible leadership on AI adoption.
- Operational changes: Those organisations that identify tasks and workflows to use AI, make appropriate changes to workflows and processes and invest in training for new skills benefit the most.
- Individual accountability: It’s critical to strike the right balance between team-wide adoption and ensuring individuals feel empowered to encourage safe innovation.
What do you do with the saved time?
The big question is what do organisations do with all the time that AI helps them save? The answer will vary depending on the organisation and its specific circumstances. They can include:
- Do more with less – use the saved time to do additional projects or serve more clients with the same number of people.
- Improve work-life balance – reduce stress and pressure by enabling employees to use the saved time flexibly. AI can definitely help employers experimenting with or adopting four-day weeks.
- Improve quality – in an increasingly complex world and competitive environment the saved time can be used to do everything better and improve performance to gain a competitive advantage.
- Freeze or reduce headcount – this is an extremely short-sighted approach as means companies simply maintain the status quo instead of using AI to grow and improve.
- Unlock innovation – or copy from Google and allow every employee time each week to invest in their own work-related innovation project.
Next steps on your AI roadmap
Companies and boards that wait for perfect clarity on AI will cede control to faster competitors. The UK’s Government Communication Service has already shown it’s possible to move at speed and at scale. Within a year, it had 3,500 of its 7,000 members using its custom-built AI assistant to help with more than 50+ communication tasks.
Communications teams need to start with a low-risk pilot this quarter to define the potential benefits and barriers that need to be overcome before rolling AI out more widely. Let us know if we can help.
