惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
云风的 BLOG
云风的 BLOG
美团技术团队
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
爱范儿
爱范儿
Stack Overflow Blog
Stack Overflow Blog
WordPress大学
WordPress大学
GbyAI
GbyAI
雷峰网
雷峰网
P
Proofpoint News Feed
IT之家
IT之家
人人都是产品经理
人人都是产品经理
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
aimingoo的专栏
aimingoo的专栏
小众软件
小众软件
T
The Blog of Author Tim Ferriss
月光博客
月光博客
V
Visual Studio Blog
L
LINUX DO - 热门话题
L
Lohrmann on Cybersecurity
T
Troy Hunt's Blog
Project Zero
Project Zero
U
Unit 42
T
Tor Project blog
Scott Helme
Scott Helme
L
LINUX DO - 最新话题
cs.CV updates on arXiv.org
cs.CV updates on arXiv.org
CTFtime.org: upcoming CTF events
CTFtime.org: upcoming CTF events
I
InfoQ
Cloudbric
Cloudbric
P
Proofpoint News Feed
The Cloudflare Blog
H
Heimdal Security Blog
Google DeepMind News
Google DeepMind News
The GitHub Blog
The GitHub Blog
Attack and Defense Labs
Attack and Defense Labs
有赞技术团队
有赞技术团队
T
Threat Research - Cisco Blogs
T
The Exploit Database - CXSecurity.com
J
Java Code Geeks
博客园 - 【当耐特】
Security Latest
Security Latest
The Register - Security
The Register - Security
F
Fortinet All Blogs
I
Intezer
H
Hackread – Cybersecurity News, Data Breaches, AI and More
MongoDB | Blog
MongoDB | Blog
N
Netflix TechBlog - Medium
NISL@THU
NISL@THU
T
Tenable Blog

The Register - Offbeat: Legal

Noyb cries foul on LinkedIn withholding profile visitor data China makes it illegal to fire humans if AI takes their jobs Databricks fails to shake authors' copyright claim Cloudera allegedly overlooked US job candidates: DoJ Australia threatens tech companies with 2.25 percent tax China blocks Meta's acquisition of AI outfit Manus Scotland Yard can keep using live facial recognition on Londoners, say judges UK tribunal sends £2B claim accusing Microsoft of overcharging for licensing to trial Yet another ex-ransomware negotiator admits turning rogue after payoff from crimelords Americans behind Nork IT fraud sentenced to 200 months Indian government investigating TCS after police sting French cops free mother and son after crypto kidnapping EFF: California 3D printer bill threatens digital freedoms IBM pays up under Trump administration's diversity blitz OpenAI CEO Sam Altman home attack suspect charged Big Tech has not enforced Australia’s social media ban Big Tech has not enforced Australia’s social media ban China's not thrilled AI experts want to leave the country China's not thrilled AI experts want to leave the country JLR cyber bailout risks dangerous precedent, watchdog warns Patel dodges question about FBI buying location data Patel dodges question about FBI buying location data ChatGPT advised exec on firing Subnautica founders: court Japan to allow ‘proactive cyber-defense’ from October 1st FSF urges AI vendors to liberate LLMs Age verification isn't sage verification when it's inside operating systems India tests whether AI can stop trains hitting elephants Perplexity Comet hurtling toward Amazon ban Lenovo, Nintendo sue US government seeking tariff refunds Google embraces third party app stores and payments OpenA says Pentagon set ‘scary precedent’ binning Anthropic China floats conspiracies about US crypto lawsuits Microsoft 'cooperating' with Japanese antitrust probe Anthropic misanthropic toward China's AI labs Americans sue Homeland Security over 'illegal' surveillance SerpApi asks court to dismiss Google web scraping lawsuit Qualcomm set to triumph in UK smartphone ‘patent tax’ case GPT-5 bests human judges in legal smack down Starlink speeds past terrestrial networks – and regulators Indian police commissioner wants ID cards for AI agents Rail workers accused of using ChatGPT for legal help Ghost gun legislation casts shadow over 3D printing UK to probe xAI over its revolting robo-smut generator UK to probe xAI over its revolting robo-smut generator Ex-Google engineer convicted of stealing AI secrets Ex-Google engineer convicted of stealing AI secrets Nudify app proliferation shows naked ambition of Apple and Google Nudify apps get past Google, Apple app moderation European Commission opens new investigation into X's Grok Meta probed over WhatsApp data disclosure Surrender as a service: Microsoft unlocks BitLocker for feds Oracle, Michael Dell, invest in JV to run TikTok USA UK gambling czar says Meta turns blind eye to illegal ads Akamai CEO wants help to defeat piracy, reckons he can handle edge AI alone Akamai CEO wants help to defeat piracy, can do edge AI alone Ofcom keeps X under the microscope despite Grok 'nudify' fix India demands crypto outfits geolocate customers, get a selfie to prove they’re real Tories vow to boot under-16s off social media and ban phones in schools Cloudflare CEO threatens to pull out of Italy Malaysia and Indonesia block X over deepfake smut EU vows to stand firm as US steps up attacks on tech regs X sues to protect Twitter brand Musk has been trying to kill Reddit sues Australia to escape kids social media ban Crypto-crasher Do Kwon jailed for 15 years Cloud group says EU should have blocked VMware-Broadcom Australia bans teens from social media – good luck with that Care leavers face bureaucracy and delays accessing records ICE-tracking app developer sues Trump administration Judge may force Vizio to share source code under GPL EU fines X €120M in first-ever DSA penalty payout IP lawyer's son surprises with vibe-coded IP infringement Campbell’s cans IT VP after ‘3D-printed chicken' rant TSMC lawsuit claims former exec probably leaks to Intel AI nudification site fined £55K for skipping age checks Senators propose to let users sue tech giants for harmful al Dutch turbine engineer tried to turn wind into crypto £5B Bitcoin bandit sent down for 11 years EU’s leaked GDPR, AI reforms slated by privacy activists Feds beat fraudster in $345M destroyed Bitcoin dispute Getty loses UK copyright battle against Stability AI Supermicro launches probe after staff charged with China export violations
AI vs the cold hard reality of the legal profession
Rupert Goodwins Rupert Goodwins · 2026-04-13 · via The Register - Offbeat: Legal

AI + ML

What happened when AI ran into the cold hard reality of the legal profession

Hallucinations don't fly in a court of law

OPINION For a sector at the heart of US economic growth, AI claims and counter-claims remain curiously hard to reconcile. Models are improving at the speed of light, AI firms claim, yet the message from the codeface remains that benefits are more than offset by the downsides.

AI can make you a 10x coder, if you spend 10x time in preparation, wrangling, and error checking. You can deploy AI agents, as long as you deploy other AI agents to watch them. AI-generated code needs AI-generated tests to cope with increased volume, and look at what that does to infrastructure stress.

Polarized opinion and commercially mandated claims have created quite the fog over the AI-coding battlefield. It's much worse in other sectors, where AI has fewer quality metrics and visible critical analyses. If only there were a well-defined area with a long history of rules-based data quality dependency, transparency, and frameworks for enforcing truth and professional standards. If that saw a sudden uptake in AI usage, we could see exactly what the tech actually does.

The good news: we have exactly that. It's called the legal system. The bad news: it's not going well.

The root cause is the same combination of two factors that go very badly together. AI is exceptionally good at producing structured documents that look as if they were generated by a human expert. AI also generates and incorporates hallucinatory facts that have the exact look and feel of reality apart from one small flaw: they're false.

This is all known well enough. The consequences are also universally acknowledged. AI's utility is very much damaged by hallucinations, and thus its output needs extensive checking. But it looks so good, so convincing, that it is very human to accept the AI promise of vastly improved productivity and hope for the best.

The result in legal systems around the world is fake cases. Lawyers making arguments in court rely on logical arguments backed up by existing case law. This comes in written depositions, where cases that agree with the argument are quoted or summarized in context, and citations provided to those cases in archives.

It was inevitable that lawyers would use AI to write depositions, prompting for the outcome they wished to prove. It was also inevitable that the AI would hallucinate cases that seemed to help the case being made. It was further inevitable that some of these would evade fact checking and end up in court. This first happened, or at least got widely noticed, in a case in the Southern District of New York in 2023. What did not seem inevitable and seems even more extraordinary is what happened next.

Lawyers have a special relationship with the legal system and with courts in particular. If they present facts, they are under strict obligation to be truthful and to have made appropriate efforts to verify what they say. They are officers of the court, and bound by professional ethics. Lawyers are also widely rumored to be human and make mistakes but they are also expected to learn from those and not repeat them. The first lawyer to put their signature on an AI-tainted deposition could plead that it was a new technology and seductively efficient, and the court may be minded to issue a stern warning and leave it at that.

Once that news got out among the legal community, a community exceptionally versed in gossiping about itself, the excuse from ignorance would not wash. You'd expect the incidence of fake cases to carry on as low-level noise, with the odd chancer trying it on, but everyone knowing the repercussions of being caught cheating by an institution jealously guarding its sanctity with effectively infinite powers of sanction.

What has happened is more akin to the early stages of plague. Six months after that first high-profile US case, another caught m'learneds' attention in a London tribunal. Last week, NPR reported that the business school HEC Paris has recorded some 1,200 cases involving hallucinations from around the world, with 800 from the US alone. Ten cases from ten different jurisdictions came in recently on the same day. The rate, they say, is still increasing.

This is despite some cases going beyond high profile, and courts everywhere ratcheting up the immune system response, fining lawyers with six figure sums. There are also proposals to require labeling of AI-generated documents, which will probably go as well as you might imagine.

The legal profession has a long tradition of making junior employees work very hard with limited resources or support from seniors. In at least one case, the underling was told to use AI to generate a brief but was not given access to the legal database they needed to check cases. Saves money, right? That the legal profession can be as exploitative as any is no surprise. That it cannot help itself but get a taste for AI that overwhelms its judgment as surely as a nose full of cocaine is seemingly indicative of how dangerous AI can be. That the problem is getting worse is also a good indication that whatever the new models do better, hallucinations ain't going away.

Responsible lawyers known to The Reg report that using AI needs as much time to verify as it saves, but that it's still worthwhile if used judiciously. This doesn't match AI hype, but it clearly matches the truth. It's also not hard to imagine mechanisms to automate case-citation checking, but there we go again on AI demanding more to do its job.

Keep a close eye on legal hallucinations. We're still in the early stages of the epidemic, and while the smart money will be on the legal system shutting down the problem before AI is fixed, it will be shut down. That will leave us with the question of how much damage AI is doing in other sectors, other organizations, where cover-your-ass behavior takes the place of professional ethics, and mutual blind eyes trump transparency and truth. Let's hope we can fix that before complacency turns into a crisis. ®