Legal-AI Updates | Fulcrum Tech Services Newsletter


A large number of articles came across my desk this week about major improvements in the Legal-AI products. It's worth checking to see what new features your Legal AI has now that you might not be aware of. The "Worth a Bookmark" is an interesting read that gives you some talking points to defend taking your time with producing legal documents rather than being rushed by the speed at which AI agents can produce drafts.


Harvey and PacerPro Announce Partnership To Connect Docket Data with AI Litigation Workflows

With this, a firm can, using Harvey, receive full context around motions and pleadings the moment they arrive. The AI would have "access to their firm’s complete litigation record, including the filings and case context from matters the firm has previously handled and those currently on the docket." This would allow the AI to summarize the posture of the case and suggest next steps immediately.


Thomson Reuters Launches Thomson, Its Own Proprietary LLM Trained on Westlaw and Practical Law Content

Most Law-AI companies use some kind of partnership with one of the big LLM labs. Thomson is built entirely independently and trained on Thomson Reuters data, which is extensive. The result may be a highly-specialized AI at lower cost than the big lab models.


LexisNexis Unveils ‘Legal Intelligence Engine,’ Rebuilding Protégé Around Dynamic Agentic Orchestration

Their new agent goes from having very scripted workflows to now being able to figure out the details of tool and model usage in the background, making it a much more natural experience. You tell it what you want it to do and it does it. I'm not really sure this is all that different from what Claude or Co-Pilot can do for you already, though.


How To Audit Your Law Firm Website For AI Readability

An introduction to the new SEO: optimization for LLM's to find your site. It discusses using structured data, and the takeaway is "talk to a web search expert." I'm not one, myself, but I can introduce you to one.


This Week's Article: Can You Answer These Five Technology Questions?

This is my new self-assessment for considering possible risks and lost opportunities. Asking yourself these questions about tech in your business might show you areas you need to improve.


Worth a Bookmark

Agentic AI And The QWERTY Problem

This is a good exploration of the use of AI by attorneys and how it affect the editing process. It offers an explanation you can share with clients as to why it is important to slow down a bit and take editing seriously.

Fulcrum Technology Services

Expertly curated news about legal tech and security. Get the latest news and information that you need to hear.

Read more from Fulcrum Technology Services

This week's theme is about AI writing. Claude adds "watermarks" to reveal what is AI written. We examine the larger implications of underhanded AI writing in legal briefs. I have an MS Word tip for collaboration and discussion in Word documents. Lastly, learn seven good rules for writing with AI. Claude’s Watermarks and their Legal Sector Impact Claude now puts "watermarks" in text it writes so people with the right tools can tell it was generated by Claude. It's based on the frequency of...

This week's collection highlights a trend I've been watching closely: legal technology is becoming increasingly specialized, practical, and accessible. LexisNexis Bets On Build-A-Bear Model For Legal AI LexisNexis will now build you a custom AI tool on the spot if you walk into their New York lab practically while you wait. This is very similar to the service we provide our clients. But, ours are even better: we come to you! As AI Legal Research Tools Race to Give Answers, This New Product...

You connect to hotel wifi without thinking twice. That's exactly what a new wave of attacks is counting on - custom malware running on the hotel's network, tricking your laptop into thinking it's talking to the real wifi, then using that foothold to steal your Microsoft 365 credentials. There's a phishing kit doing something similar right now by spoofing RingCentral notifications - both stories are in this week's roundup below. Most firms have their Microsoft 365 locked down about as tight as...