The future of the patent profession: Are we looking into an AI abyss?

AI presents a huge dilemma for patent attorneys. There is no doubt that AI will have a dramatic impact on the profession and the business model that many firms have relied on for decades. Faced with this challenge, there is a growing divide in the profession between the firms and attorneys who are actively adopting AI, and those that remain stubbornly AI-sceptic. Amongst the AI adopters, there is also a huge diversity in marketing approach. Whilst some firms are choosing not to mention their use of AI at all on their websites, others are actively marketing AI-based discounts. Where does all this leave the profession? 

Heads in the sand

The abyss 

Many patent attorneys remain highly AI-sceptic. Indeed, many attorneys openly admit to using no AI at all in their professional work. This admission is usually followed with an anecdote about an AI they used a year ago that made a hilariously-incorrect hallucinatory error, or a story about some terrible AI-drafted claims. 

The hesitation felt by many in the profession is logical if you consider that much of the patent profession relies heavily on the drafting and prosecution billable hour. If a firm adopts an AI tool that significantly reduces the time required to, for example, draft a prosecution response, it faces a direct threat to its revenue stream. If the volume of work remains static and is dictated by the frequency of EPO office actions, then speeding up the process simply erodes revenue.  It is therefore no surprise that large firms focused on high volume drafting and prosecution seem less than enthused by AI. 

However, simply ignoring AI is unlikely to be sustainable. With AI capabilities ever increasing, cost-conscious clients are increasingly likely to demand at least some AI-assisted efficiency that translates into reduced fees. 

The race to the bottom

In contrast to the AI-naysayers and sceptics, some firms are actively marketing cheaper services based on substantial AI-assisted reductions in patent prosecution costs. This is not just start-ups and AI-native law companies, but Tier 1 firms. Beyond the high-level marketing-spiel, it remains unclear how such reductions (up to 30% in some cases) are calculated. If the same level of quality is promised from the AI-assisted service, this approach would also seem to be kickstarting a race to the bottom. There may also come a point where clients may simply ask why they are paying their attorney to ask ChatGPT to draft and prosecute their patent, when they could simply do this for themselves.

There is a mode of thought that AI will kill the billable hour, replacing it with a fixed-fee model. However, shifting to fixed fees does not solve the problem, given that it may still be difficult to convince clients that they should be paying high fees when the attorney is using off-the-shelf AI tools to complete the same work as before in a fraction of the time. The firms refusing to adopt AI may then find themselves out-competed by those who are. 

AI for AI's sake

In the middle, are all the firms that have rolled out AI tools firm-wide but without a clear external narrative, with their websites often remaining silent on the use of AI or providing only vague references. It seems that firms are struggling to define exactly what problem they are solving for their clients or their business. The internal AI-task-force of attorneys tasked with trialling off-the-shelf tools for the firm appears to be almost a universal.

However, firms simply seeking to adopt an off-the-shelf AI-wrapper (simply because they feel they have to) are putting themselves at the mercy of a highly fluid and precarious legal tech market (IPKat). These firms are at risk of becoming overly reliant on specific service providers whose products may quickly become obsolete. We have already seen this pattern in the software industry where specialized AI coding solutions, such as Devin AI, were rapidly superseded by the native capabilities of LLMs like Claude Code, despite huge initial investment and venture capital excitement. As the Vice President of Google’s Global Start-up Organization Darren Mowry put it, "if you're really just counting on the back-end model to do all the work and you're almost white-labelling that model, the industry doesn't have a lot of patience for that anymore". As AI models improve the functionality gap that wrappers once filled is closed by the model providers themselves at dramatically reduced costs. We are already seeing this in the legal tech world with the recent release of Claude Legal.

Patent firms that invested heavily in complex legal tech platforms twelve months ago may find themselves tethered to expensive and unnecessary AI wrappers that might not survive the next two years. Firms must therefore make a conscious choice about the value proposition of their chosen tools rather than simply adopting technology because they feel they must.

Final thoughts

Learning how to use AI is one thing, understanding why you should use AI is another. Whilst there is certainly currently a lot of AI hype, it is also certain that AI will change the patent profession. We are an industry based on the digestion, interpretation and production of the written word, all tasks that AI excels at. The pace of change is also astounding. The profession is inevitably going to look very different in 10 years' time from what it does today. Patent attorneys are therefore going to need to figure out not only how to use AI, but why they should use it.

Further reading 

• The spectre of hallucination
• Solving the confidentiality problem
• Are you behind the wheel or waiting for the bus?