Recent university graduate Lotte Underwood is a public relations executive with a twist – she spends more time using AI than actually doing PR. Underwood, 25, joined London-based PR agency Fight or Flight in November last year, in the new role of “AI associate” after completing a degree and then a masters in marketing. She’s part of a recently established, dedicated AI team that has built around 30 AI agents. Specific tasks include industry research, client reporting, media interview preparation, helping clients create clearer briefs, and image generation.
A big part of Underwood’s job is to manage those agents, check their outputs and feed those to the team members who are dealing directly with clients and the media. Another key responsibility of hers is to monitor how clients appear in the responses from AI tools like ChatGPT, Claude, Gemini and others. “A huge part is being in client meetings and calls to work out what is actually going to be most helpful from these agents,” says Underwood. What is an AI agent?
AI is seen as one of the main factors behind graduate job cuts, but companies are also creating dedicated AI-specific graduate roles like Underwood’s. Dillan Patel, for example, joined engineering software company Stride 18 months ago as an intern. A business management graduate, he was put to work developing an internal chatbot designed to capture the knowledge of a senior staff member transitioning into retirement. Patel, 24, is now an AI automation analyst, tasked with developing further internal agents to provide specific information to individual teams, while also identifying which manual or repetitive tasks across the company could be done by AI.
“My role focuses on finding practical ways to use technology, automation, data and AI to improve how we operate,” says Coventry-based Patel. Such AI savvy graduates are also in demand from companies like ArvatoConnect, an outsourced contact centre company. CEO Debra Maxwell says the company’s usual graduate intake is on hold this year, but she plans to resume it next year with a bigger focus on AI. “We would be looking for more AI-disruptor type roles, where you can challenge us as an organisation to think more holistically about where AI can be applied and where we could grow our business from,” says London-based Maxwell.
“We see our graduates as not filling administrative roles, but actually adding value to the organisation in a different way.” Maxwell adds: “For example, one of our recent graduates is using AI to do deep dives on specific markets, and provide more information for when we’re bidding [for new contracts]. It’s basically giving him a more rounded role.” If AI can make a graduate job bigger and more integral to the organisation, Maxwell argues that companies can create a more loyal, engaged workforce that is able to contribute more meaningfully, much earlier. Research from consultancy PwC found that some graduate roles are assuming more responsibilities because of AI, as it allows companies to demand more senior abilities from junior hires. For the consultancy’s 2026 AI Jobs Barometer, external, it analysed 2.4 million entry-level jobs in the US.
It found that the entry-level roles most influenced by AI are now seven times more likely to require traditionally senior-level skills. These include motivational leadership, strategic decision making and team building. Those entry-level roles with higher expectations have in fact grown 35% since 2019, PwC found, while other entry-level roles actually shrank by 10%. Cian Clinton, who joined consultancy Valliance last September after graduating from a Bachelor of Science in 2024, says the incorporation of AI into his role makes him feel less like a graduate, and more like an equal team member.
He uses around 10 AI agents daily for different tasks. “Your ideas and your creativity are a lot more valued because you actually get to implement those firsthand, and the responsibility you take on is a lot greater because of that,” says London-based Clinton, 23. “I am now taking on responsibilities which would usually take me four or five years to get to.” His career progression, he believes, will now be faster as a result. “AI is killing the ‘pyramid’ approach…progression is no longer based on your knowledge bank.
Now it’s more based on your critical thinking and understanding of the processes,” says Clinton. There’s no doubt Underwood, Patel and Clinton are developing valuable AI knowledge which can help them across their careers. But with their time at work so focused on AI, does it matter how much – or how little – time they’re spending learning about the professional field they’re supposed to be operating in? Underwood says she’d like to still pick up the traditional PR skills her role isn’t currently focused on, but is confident of her future in the agency’s growing AI team.
“There’s certain things that I haven’t had the chance to do yet, like directly pitching to the media. I absolutely plan to do this,” she says. Graduates becoming too AI-centric is an issue for Simon Pratt, chief operating officer at Portman Finance Group. He says being able to use AI effectively is important, but the company’s two-year graduate programme is fully centred on “learning the trade”.
“Our graduates rotate across four areas at six-month intervals, working directly alongside senior relationship managers to learn their trade,” Northampton-based Pratt explains. “They’ll learn all of our financial products, how to draft credit papers for submission to lenders, and how to build meaningful relationships with our customers. They’ll progress to running transactions from start to finish.” Claire Reed, PwC UK’s chief technology and innovation officer, believes graduate roles that are too AI heavy won’t necessarily provide graduates long-term career progression. “There will be no graduate job now that doesn’t require a baseline of AI literacy,” says Glasgow-based Reed.
“But AI skills on their own are not going to help you in your career. You’ve got to couple that with that sector knowledge of what that client actually wants.” Back in Coventry, Dillan Patel is confident his AI work will continue to be valuable, and he wants to keep building on it. “I don’t think I’d want to explore different positions within the business. Through my involvement in AI-related projects, I already work closely with different areas of the business and I’m gaining valuable knowledge and experience along the way,” he says.
Why are more roofs not made of solar tiles? X-planes: Are they needed in the new era of drones? The virtual worlds where robots are trained
