ARTIFICIAL INTELLIGENCE HALLUCINATIONS – CAUSING PROBLEMS FOR THE PROFESSION AGAIN: “… THIS DECISION ILLUSTRATES THE CONSIDERABLE PROCEDURAL DIFFICULTIES WHICH CAN ARISE WHEN INAPPROPRIATE USE IS MADE OF AI”
I have today received a copy of the second edition of Andrew Hogan’s book on AI for Lawyers (a review of which will be appearing soon). Problems with the use of AI continue to blight the profession (at home and…
WHAT HAS CIVIL LITIGATION BRIEF DONE ABOUT … THE “AUTHENTICITY” OF DOCUMENTS (OR OTHERWISE)? TEN KEY POINTS FOR PRACTITIONERS CONSIDERING THE RISKS POSED BY ARTIFICIAL INTELLIGENCE
The post earlier this morning where the Upper Tribunal considered the problems with very authentic looking documents being produced by Artificial Intelligence is illustrative of a very real problem for civil practitioners. The judgment contains a timely reminder that documents…
FALSE DOCUMENTS, LATE EVIDENCE AND ARTIFICIAL INTELLIGENCE: A WHOLE HOST OF ISSUES THAT MAKE JUDICIAL LIFE MORE DIFFICULT…
Much of the discussion in relation to Artificial Intelligence to date has related to its ability to “hallucinate” false cases. However it can also be used to create extremely realistic false documents. This is an issue considered by the judges…
AI AND CIVIL JUSTICE: PREPARING FOR THE TSUNAMI – LORD BRIGGS ON ACCESS TO JUSTICE, PROCEDURAL RULES AND THE HUMAN JUDGE
Issues relating to the use of AI continue to “interest” the legal profession. They also interest judges. Here I summarise Lord Briggs’ Keynote Address to the Oxford Civil Justice Systems in the 21st Century Conference. “… if human judges want…
WHEN CHAT GPT DRAFTED A “SKELETON” ARGUMENT – 300 PAGES AND 132,000 WORDS LONG: THE LITIGANT RETAINS RESPONSIBILITY AND MUST COMPLY WITH THE RULES
We have here another example, in a judgment today, of the difficulties caused by the use of AI. ChatGPT “helpfully” drafted a “skeleton” argument that was 300 pages long (the maximum page length allowed was 20 pages). This was coupled…
LAWYERS AND ARTIFICIAL INTELLIGENCE AVOIDING THE PITFALLS (5): “THE HEARING WAS A SALUTARY LESSON IN HOW AI MAY PROMPT LAY PERSONS TO SAY THE RIGHT THINGS IN PLEADINGS, BUT CANNOT FURNISH THE EVIDENCE NEEDED TO ESTABLISH THOSE STATEMENTS”
This morning we are travelling to Scotland to look at a case where the judge indicated that the use of AI had led the pursuers (claimants) to bring forward a claim that was largely unsupported by evidence. This led to…
LAWYERS AND ARTIFICIAL INTELLIGENCE AVOIDING THE PITFALLS (4): EXPERTS AND THE USE OF AI: A LOT OF DANGER AREAS HERE: 20 KEY POINTS
Here we consider the risks that occur when experts use AI when they prepare a report or give evidence. We have seen several examples on this blog of evidence going awry. Here we draw upon guidance from previous posts, but…
LAWYERS AND ARTIFICIAL INTELLIGENCE AVOIDING THE PITFALLS (3): AI AND WITNESS EVIDENCE: VERY DANGEROUS BEDFELLOWS: TWENTY KEY POINTS
We have looked several times at the dangers of using Artificial Intelligence to draft witness statements. It would be naive to believe that this issue is not going to arise again. Indeed it may well already be a common feature…
LAWYERS AND ARTIFICIAL INTELLIGENCE AVOIDING THE PITFALLS (2): HOW TO SPOT “FALSE” AUTHORITIES AND WHAT TO DO ABOUT IT (PLUS SOME USEFUL LINKS)
Many practising lawyers will now have seen cases where AI has been used to put forward legal propositions. Sometimes this involves “hallucinated” cases, statutes or rules. More problematically a real case or statute is cited, however the citation is wrong…
BACK TO BASICS MONDAY: LAWYERS AND ARTIFICIAL INTELLIGENCE (1): THREE IMPORTANT PRINCIPLES: YOU OWN THE WORK; AI IS NOT AN AUTHORITATIVE SOURCE OF LAW; IT IS NOT ENOUGH THAT THE CASE EXISTS…
The post earlier today on the misuse of AI when defending an SDT hearing has led me to start a short series about the dangers of lawyers (litigators in particular) using AI. This is not an anti-AI stance. There is…
MORE ON THE USE OF ARTIFICIAL INTELLIGENCE AND “HALLUCINATING” CASE LAW AND LEGAL PRINCIPLES: THE SOLICITORS DISCIPLINARY TRIBUNAL GIVES A CLEAR VIEW…
We have now looked at numerous court cases where lawyers have come to grief because of the use of Artificial Intelligence to cite law incorrectly, or simply manufacture false cases and equally false legal principles and citations. Here we have…
ARTIFICIAL INTELLIGENCE IN COURT: AN INTERESTING PRACTICE DIRECTION FROM THE HIGH COURT IN IRELAND : GOOD ADVICE FOR US ALL
In England and Wales we are still considering appropriate guidance for the use of Artificial Intelligence in the courts. The High Court in Ireland has issued a specific Practice Direction on the use of Generative AI in Court Documents. It…
WITNESS EVIDENCE WEDNESDAY (2): WHEN A WITNESS GETS COACHING FROM ARTIFICIAL INTELLIGENCE – WHAT ARE THE IMPLICATIONS?
Artificial Intelligence continues to give rise to novel legal issues. In this case the courts had to consider the impact of clear evidence that the complainant in a criminal case had received detailed “coaching” by the use of A.I. …
HALLUCINATED CASE LAW AND THE CROWN PROSECUTION SERVICE (OF ALL BODIES…)
All cases of hallucinated cases are serious, however some may be more serious than others. If we have a situation where the Crown Prosecution Service files documents which rely on cases that simply do not exist, this has to be…
THE TCC GUIDANCE ON THE USE OF ARTIFICIAL INTELLIGENCE: A DETAILED EXAMINATION
The section on the use of AI in the latest edition of The Technology and Construction Court Guide is relatively short. However it is significant. It shows that the courts recognise that AI is going to be used. “All legal…
ARTIFICIAL INTELLIGENCE AND LITIGATION: AN UPDATE ON CONSULTATION FINDINGS FROM THE CIVIL JUSTICE COUNCIL
The Civil Justice Council is carrying out a Consultation on the Use of AI for the preparation of court documents. It published an update on its consultation findings yesterday. The preliminary view appears to be that the use of AI…
WHAT HAPPENS WHEN A PARTY DOES NOT RESPOND TO A NOTICE TO ADMIT FACTS (THERE IS NO “DEEMED ADMISSION”) PLUS ANOTHER POSSIBLE “HALLUCINATED” “FICTITIOUS” RULE
This case has two important practical points. Firstly (contrary to the case put forward by the claimant) a failure to respond to a Notice to Admit facts does not give rise to an “implied admission” by the recipient of the…
THE USE OF ARTIFICIAL INTELLIGENCE – LANDING SOLICITORS IN SERIOUS TROUBLE AGAIN (AND DON’T BLAME THE AI FOR EVERYTHING – IT ACTUALLY GAVE OUT WARNINGS TO CHECK…)
One day the incorrect use of AI to cite “hallucinated” authorities is going to ruin someone’s career. It may have done so already, there are a number of SRA investigations pending. The example we look at here is highly educational…
ARTIFICIAL INTELLIGENCE AND THE CITATION OF MISLEADING AUTHORITIES: ANOTHER WEEK, ANOTHER CASE: IF YOUR NAME IS ON THE DOCUMENT YOU “OWN” IT…
We are looking at another case where the judge has expressed major concerns about the use of Artificial Intelligence in the preparation of documents for the court. The situation is now a (depressingly) familiar one where the use of AI…
MORE USE OF AI: MORE HALLUCINATED CASES: THERE IS “NO PROBLEM” IN USING AI: BUT CONSIDERABLE CARE HAS TO BE TAKEN
If these issues continue as they have been we may soon be seeing an “AI Tuesday” to add to the other themes we examine throughout the week. Here we look at another “hallucination” case which ended with the person involved…
GUIDANCE FOR THOSE USING ARTIFICIAL INTELLIGENCE TO CARRY OUT LEGAL RESEARCH: THERE IS AN OBLIGATION NOT TO ADVANCE SUBMISSIONS BASED ON “FAKE” AUTHORITIES…
There appears to be many hundreds (possibly thousands) of cases throughout the world where litigants (and often their lawyers) have relied on “hallucinated” cases, or real cases which do not, in fact, contain the quotations relied on or support the…
HALLUCINATIONS KEEP APPEARING IN THE REPORTS: TWO MORE EXAMPLES: COUNSEL AT FAULT IN BOTH…
We are looking at two more examples of “hallucinated” cases appearing in reported cases. In both cases it was counsel that was presenting the case. (In one case counsel was acting for himself). “The incident does, however, demonstrate vividly the…
WITNESS EVIDENCE WEDNESDAY: THE USE OF ARTIFICIAL INTELLIGENCE IN THE DRAFTING OF WITNESS STATEMENTS: “IT IS DIFFICULT TO DISTINGUISH BETWEEN WHAT F SAYS AND WHAT AN ALGORITHM TELLS F TO SAY”
There is much material about witness evidence and witness statements on this site. In recent years we have also been discussing the use (and misuse) of artificial intelligence. We can be fairly sure that there will be much more about…
A FURTHER EXAMPLE OF ARTIFICIAL INTELLIGENCE GENERATING PHANTOM REFERENCES AND FALSE QUOTATIONS
We see the another example of the dangers of the use of Artificial Intelligence in this case. Two authorities relied upon by a respondent did not contain the words attributed to them, none of them supported the propositions that had…
PROVIDING LEGAL SUBMISSIONS WITH INACCURATE CASE SUMMARIES: THE REPRESENTATIVE WHO WOULD “NEITHER CONFIRM OR DENY” THAT AI WAS USED
We are returning to the vexed issue of the (mis) use of Artificial Intelligence when providing written submissions to the court (in the case the First Tier Tribunal). The judge found that summaries provided were inaccurate. The judgment points…
REVIEW OF THE YEAR (2) – THE WORST PART: “HALLUCINATED” CASES IN THE COURTS: HOW IS ARTIFICIAL INTELLIGENCE BEING HANDLED & WHAT DOES THE FUTURE HOLD FOR AI AND LAWYERS?
If I had to pick the most frightening development of the year it is the revelation that parts of the legal profession have been relying on “hallucinated” (that is false) cases they have “found” by using Artificial Intelligence. Here we…
AI USED IN THE WRITING OF A JUDICIAL DECISION: READ ALL ABOUT IT…
There has been much discussion of the advantages, and disadvantages, in lawyers using AI. This is clearly going to be a major issue for the legal profession going forwarded. Last week I reviewed Andrew Hogan’s book on this topic. There…
You must be logged in to post a comment.