This fall, in-house IP teams are facing a familiar ritual: pull last year's numbers, brace for a budget cut, and figure out what to trim without gutting the portfolio's real value. On Tuesday, I participated in a IPWatchdog webinar with Gene Quinn and Jason Harrier (AI Patent Counsel, Salesforce) that tackled this exact problem: how AI is reshaping the budgeting and cost-cutting process for 2027, and what teams of any size can do about it right now.
[Watch the Webinar Recording here]
Here are the key themes and takeaways for in-house counsel and the outside firms who support them.
The budget squeeze in-house teams felt last year is still here. Live polling of the audience backed this up in blunt numbers: 44% of in-house counsel said they're facing significant pressure to cut costs from above, and another 42% said the pressure is present but manageable. Only 3% said they're feeling no pressure at all.
Outside counsel are feeling the same squeeze from the other direction. When asked whether clients are pushing them to cut fees, 46% said yes from most clients, and another 33% said yes from at least a few. Only 6% said they haven't felt any pressure yet.
Put those two together and the picture is clear: cost pressure is now the default operating condition on both sides of the table, not an occasional headwind.
What's different from last year is where the cuts are coming from. Last year's low-hanging fruit, like 11.5 year maintenance fee pruning and outside counsel consolidation, has largely been picked. This year, teams are looking harder at filing counts (especially in foreign jurisdictions), continuation strategy, and how to optimize their remaining OC firm roster by technology expertise.
Gene's take: filing less is the single biggest cost lever available, since every downstream cost (drafting, prosecution, maintenance fees) disappears if the application never gets filed. But the panel was equally clear that "file less" only works if it's paired with "file better."
AI is now enabling teams to:
Jason described this as being able to "spearfish" the best innovation across a large organization instead of reactively filing whatever inventors happen to remember. The upshot for smaller teams: this isn't a scale-dependent capability. A team of one to five people can build the same kind of scoring discipline as a 5,000-patent portfolio, just at a smaller volume.
Once a case is in prosecution, the panel's advice was to stop reviewing every matter equally and start surfacing the ones that need attention. AI can now flag:
Both Jason and Gene made the same point from different angles: outside counsel rarely recommend killing a case on their own, and in-house counsel have historically avoided the decision too, partly for defensible-record reasons and partly because abandoning something feels like admitting a mistake. Data gives counsel the objective backup to make it and defend it. As Jason put it, a CFO is never going to be upset about saved money, provided you can explain the risk-and-reward math clearly.
I walked through how outside counsel scorecards have gotten more sophisticated. Instead of relying on general allowance rates, teams are now combining:
Gene's caution here is worth repeating for firms: numbers alone don't tell the whole story. A firm might rank lower because it inherited a difficult transfer portfolio or works in an inherently tough technology area. Firms should know their own numbers before a client's spreadsheet surprises them, and should have a clear, ready explanation when the data doesn't tell the full story.
Maintenance fees came up as the audience's top area of interest, and the panel's guidance centered on building an internal value-scoring system rather than treating every fee decision the same way. Useful inputs include citation counts, whether an asset is a continuation, licensing activity, and increasingly, AI-driven analysis of whether a patent's claims actually map to something a competitor is doing in the market. That last point stood out: general-purpose AI models are weak at reasoning over patent data alone, but they're very strong at connecting patent claims to publicly available competitor information, which is often the clearest signal of whether an asset is worth the next fee payment. Combine that with Juristat’s MCP server for curated patent data in your AI tools, and you have a really robust analysis.
A recurring question was how this applies to leaner teams, for example a one to five person department managing a smaller budget. The panel's answer was reassuring: smaller teams may actually be better positioned, since they can build scoring and forecasting habits from scratch rather than retrofitting them onto a massive legacy portfolio. The advice was consistent across both panelists: pick one workflow, whether that's maintenance fee review or a whitespace analysis, and start there. The tools are improving quickly enough that a modest starting point tends to expand in capability faster than expected.
One of the most concrete shifts discussed was budget forecasting itself. Where teams used to spend weeks in spreadsheets, or rely on rough assumptions like "every application gets two office actions," AI models can now project remaining prosecution activity for in-flight applications based on examiner and art unit history, layer in actual flat fees and USPTO fees, estimate future filing volume from historical run rate, and factor in expected maintenance fee pruning, all in a single day when the underlying data is connected properly.
A final live poll asked in-house attendees directly where they expect to cut costs most in the year ahead. Maintenance fee pruning topped the list at 39%, followed by more prosecution efficiency at 26%, outside counsel fee reductions or consolidation at 20%, and filing less at 16%.
Gene pushed back on this result in real time, and it's worth flagging for readers too. His view: filing less should arguably be the biggest lever, not the smallest, since every downstream cost (drafting, prosecution, maintenance fees) disappears entirely when an application never gets filed in the first place. The audience's ranking suggests most teams are still more comfortable trimming an existing portfolio than tightening the filing gate itself. That gap between where teams expect to find savings and where the panel thinks the biggest savings actually live is worth a second look inside any budgeting conversation this cycle.
Nothing about the underlying budgeting exercise has changed. Teams still need to file smart, manage prosecution efficiently, evaluate outside counsel fairly, and prune maintenance fees strategically. What's changed is the ability to do all of it with real data rather than gut instinct, at a pace and depth that wasn't practical a year ago. For in-house teams, the recommendation is to pick one workflow and build from there. For outside counsel, the opportunity is to bring this kind of analysis to clients proactively, turning cost pressure into a chance to demonstrate value rather than just absorb rate pressure.