theguardian

Studios accuse AI firm of ‘piracy’ and seek injunction over alleged use of copyrighted characters

DisneyandUniversalsued anartificial intelligencecompany on Wednesday, alleging copyright infringement. In their lawsuit, the entertainment giants called Midjourney’s popular AI-powered image generator a “bottomless pit of plagiarism” for its alleged reproductions of the studios’ best-known characters.

The suit, filed in federal court in Los Angeles, claims Midjourney pirated the libraries of the two Hollywood studios, making and distributing without permission “innumerable” copies of their marquee characters such as Darth Vader from Star Wars, Elsa from Frozen, and the Minions from Despicable Me. Midjourney did not immediately respond to a request for comment.

The suit by Disney and Universal over images and video represents a new frontier in the raging legal wars over the copyright and the creation of generative artificial intelligence. Previous suits have covered copyrighted text and music; Disney and Universal are two of the biggest industry players thus far to sue over images and videos.

Horacio Gutierrez, Disney’s chief legal officer, said in a statement: “We are bullish on the promise of AI technology and optimistic about how it can be used responsibly as a tool to further human creativity, but piracy is piracy, and the fact that it’s done by an AI company does not make it any less infringing.”

NBCUniversal’s executive vice-president and general counsel, Kim Harris, said the company was suing to “protect the hard work of all the artists whose work entertains and inspires us and the significant investment we make in our content”.

The studios claim the San Francisco company, one of the earliest to offer an AI product that generated images, rebuffed their request to stop infringing on their copyrighted works or, at a minimum, take technological measures to halt the creation of AI-generated images that copied the characters.

Instead, the studios argue, Midjourney continued to release new versions of its AI image service that boast higher-quality infringing images. Midjourney can recreate animated images from a typed request, or prompt. AI companies use enormous troves of data, often retrieved from millions of websites, to train programs to generate new images or text.

In a 2022 interview with Forbes, Midjourney CEO Holz said he built the company’s database by performing “a big scrape of the Internet”.

In the suit filed by seven corporate entities at the studios that own or control copyrights for the various Disney andUniversal Picturesfilm units, the studios offered examples of Midjourney animations that include Disney characters, such as Yoda wielding a lightsaber, Bart Simpson riding a skateboard, Marvel’s Iron Man soaring above the clouds and Pixar’s Buzz Lightyear taking flight. The image generator also recreated such Universal characters as How to Train Your Dragon’s dragon, Toothless, the green ogre Shrek and Po from Kung Fu Panda.

“By helping itself to plaintiffs’ copyrighted works, and then distributing images (and soon videos) that blatantly incorporate and copy Disney’s and Universal’s famous characters – without investing a penny in their creation – Midjourney is the quintessential copyright free-rider and a bottomless pit of plagiarism,” the suit alleges.

Disney and Universal asked the court for a preliminary injunction, to prevent Midjourney from copying their works, or offering its image- or video-generation service without protections against infringement. The studios also seek unspecified damages.

Midjourney, founded in 2021 by David Holz, monetizes the service through paid subscriptions and generated $300m in revenue last year alone, the studios said.

A weekly dive in to how technology is shaping our lives

This is not the first time Midjourney has been accused of misusing artists’ work to train their AI systems. A year ago, a California federal judge found that 10 artists behind a copyright infringement suit against Midjourney, Stability AI and other companies had plausibly argued these AI companies had copied and stored their work on company servers, and could be liable for using it without permission. That ruling allowed the lawsuit over the unauthorized use of images to proceed. It is in the process of litigation.

The cases are part of a wave of lawsuits brought by copyright owners including authors, news outlets and record labels against tech companies over their use of copyrighted materials for AI training without permission.

Asked whether he sought consent of the artists whose work was covered by copyright, Holz responded: “there isn’t really a way to get a hundred million images and know where they’re coming from.” OpenAI said early last year in a filing to the UK government that it would be “impossible to train today’s leading AI models without using copyrighted materials”.

In late 2023, the New York Times sued OpenAI, maker of ChatGPT, and Microsoft, which owns a 49% stake in the startup, alleging unauthorized use and reproduction of its articles, which comprise a vast corpus of text. The suit is ongoing. Other news outlets, including the Guardian, have reached agreements to license their archives to artificial intelligence companies. Book authors have likewise sued Meta for alleged use of a huge database of pirated books to train its Llama AI models, though many of the writers’ claims have been dismissed.

In June 2024, major record labels sued two AI companies for copyright infringement. Sony Music Entertainment, Universal Music Group Recordings and Warner Records accused Suno and Udio of exploiting millions of songs to create engines that could “spit out” obviously derivative music.

By slashing time and cost of restoration, technique could be used on paintings not valuable enough for traditional approach

The centuries can leave their mark on oil paintings as wear and tear and natural ageing produce cracks, discoloration and patches where pieces of pigment have flaked off.

Repairing the damage can take conservators years, so the effort is reserved for the most valuable works, but a fresh approach promises to transform the process by restoring aged artworks in hours.

The technique draws on artificial intelligence and other computer tools to create a digital reconstruction of the damaged painting. This is then printed on to a transparent polymer sheet that is carefully laid over the work.

To demonstrate the technique, Alex Kachkine, a graduate researcher at Massachusetts Institute of Technology, restored a damaged oil-on-panel work attributed to the Master of the Prado Adoration, a Dutch painter whose name has been lost, as a late 15th-century painting after Martin Schongauer.

The painting is extremely detailed but visibly split into four panels, covered in fine cracks and dotted with thousands of tiny patches where paint has fallen off.

“A lot of the damage is to small, intricate features,” said Kachkine, who estimated it would have taken about 200 hours to restore the painting with traditional conservation techniques. “It has undergone centuries of degradation.”

Kachkine started with a scan of the painting to determine the size, shape and position of the damaged areas. This identified 5,612 separate sections that needed repair.

A digital mask was then constructed in Adobe Photoshop. To restore missing specks of paint, spots were added and colour-matched to surrounding pigments. Damage to patterned areas was corrected by copying similar patterns from elsewhere in the painting. The missing face of an infant was copied from another work by the same artist.

Once finished, the mask was printed on to a polymer sheet, varnished to prevent the ink from running and overlaid on the painting.

In all, 57,314 colours were used to infill damaged areas. The corrections are designed to improve the painting, even if they are not perfectly aligned.

On seeing the result, Kachkine was delighted. “It followed years of effort to try to get the method working,” he said. “There was a fair bit of relief that finally this method was able to reconstruct and stitch together the surviving parts of the painting.”

The approach,described in Nature, can only be used on varnished paintings that are smooth enough for the sheet to lie flat on. The mask can be peeled off or removed using conservators’ solvents, leaving no traces on the original artwork.

Kachkine hopes the method will allow galleries to restore and display scores of damaged paintings that are not deemed valuable enough to warrant traditional restoration.

But he acknowledges there are ethical issues to consider, such as whether having a film covering a painting is acceptable, whether it hampers the viewing experience, and whether particular corrections, such as copied features, are appropriate.

In an accompanying article, Prof Hartmut Kutzke at the University of Oslo’s Museum of Cultural History, said the approach provided a way to restore damaged paintings faster and more cheaply than was possible with conventional techniques.

“The method is likely to be most applicable to paintings of relatively low value that would otherwise be housed behind closed doors, and might not be suitable for famous, valuable artworks,” he said. “However, it could widen public access to art, bringing damaged paintings out of storage and in front of a new audience.”

Tim Ayres says the Albanese government will focus on legislation and regulation but country would benefit from moving quickly

Australia news live: latest politics updates

Get ourbreaking news email,free appordaily news podcast

Australia must “lean in hard” to the benefits of artificial intelligence or else risk ending up “on the end of somebody else’s supply chain”, according to the new industry and science minister, Tim Ayres, with the Labor government planning to further regulate the rapidly evolving technology.

Ayres, a former official with the manufacturing union, acknowledged Australians remained sceptical about AI and stressed that employers and employees needed to have discussions about how automation could affect workplaces.

The minister said Australia had “no alternative” but to embrace the new technology and seek to become a world leader in regulating and using AI.

“It’s the government’s job to lean into the opportunity to outline that for businesses and for workers, but also to make sure that they are confident that we’ve got the capability to deal with the potential pitfalls,” Ayres told Guardian Australia.

“I think the Australian answer has got to be leaning in hard and focusing on strategy and regulation that is in the interest of Australians.”

Ayres was promoted to cabinet last month after holding the junior portfolio of assistant minister for manufacturing and trade last term.

Sign up for Guardian Australia’s breaking news email

Now as minister for industry, innovation and science, Ayres has greater direction of the Labor government’s flagship Future Made In Australia initiative – the umbrella plan tying together policies around manufacturing, the energy transition, research and business.

One of Ayres’ more public-facing and immediate challenges, however, is around AI policy. His predecessor, Ed Husic, set up several processes around the development of local industry and mandatory guardrails around its use, including discussion of a new standalone AI act and legislation.

Ayres, who has been minister for less than a month, says the government is still setting its course of action, considering approaches from like-minded countries and the breakneck speed of the technology’s evolution. He said the response would include legislation and regulation, which was still to be decided, but that Australia would benefit from moving quickly.

“There is no alternative to the approach of Australia leaning in,” he said. “That’s how we’re going to shape a digital future for Australia. That’s how we get some agency and say over the way that the technology develops and our global partners develop on these questions.

“The alternative is to sit back right and just to be on the end of somebody else’s supply chain.”

The minister said there would be “a lot to gain” from embracing AI, particularly highlighting potential boosts to productivity and economic growth. Ayres, who grew up on a cattle farm near Lismore, said both white-collar and traditional blue-collar jobs could reap benefits from adopting automation and new technology.

He said that his experience in the manufacturing union was that many workers accepted “the cold reality that the only thing more disruptive in the labour market is standing back and allowing Australia to become a cul-de-sac in technological terms”.

“But I do want to see firms and employers being thoughtful about making sure that we’re adopting AI in a way that makes jobs better,” Ayres said.

“We’ve got an industrial relations framework that provides for proper consultation and engagement at the firm level … there’s going to be discussions workplace by workplace, around these questions.”

Last week, the Business Council of Australiareleased a major reportcalling for Australia to become a global leader in AI, talking up its potential to boost productivity and lift living standards through economic growth.

The Australian Council of Trade Unions in Decemberwarned one in three Australian workers are at risk of job lossfrom the introduction of AI.

“The BCA paper released last week pointed to a high level of scepticism amongst Australians about this new wave of technology. That’s not unusual for Australia,” Ayres said.

“Every wave of technological change reshapes the labour market. That’s the truth of it. The negative consequences in jobs terms of technological change have historically been outweighed by the new investment and the new developments in jobs and technologies.”

Ayres also said Labor would continue with its Future Made In Australia agenda, which would include an “aggressive” focus on critical minerals, iron and steel production and boosting manufacturing as part of the renewable energy transition.

“I want to make sure that we’re doing everything we can to be building new factories and industrial capability,” he said.

“Particularly in regions like central Queensland and the Hunter and La Trobe, where the intersection of our future energy advantage and our industrial capability really means Australia can deliver, and deliver for those communities as well.”

Citizen Lab says it found ‘digital fingerprints’ of military-grade spyware that Italy has admitted using against activists

The hacking mystery roiling the Italian prime minister Giorgia Meloni’s rightwing government is deepening after researchers said they had found new evidence that two more journalists were targeted using the same military-grade spyware that Italy has admitted to using against activists.

A parliamentary committee overseeing intelligence confirmed earlier this month thatItalyhad used mercenary spyware made by Israel-based Paragon Solutions against two Italian activists.

But the same committee, which launched an investigation into the hacking scandal in March, said it had been unable to determine who was behind the targeting of a prominent Italian investigative journalist, Francesco Cancellato, whose news outlet has been critical of the Meloni government.

Now areport by researchers at the Citizen Labhas revealed that a close colleague of Cancellato, Ciro Pellegrino, who is head of the investigative outlet Fanpage.it’s Naples bureau, was also targeted by a user of Paragon Solution’s spyware, which is called Graphite. The Citizen Lab said a third journalist, who chose to remain anonymous and is described as a “prominent European journalist”, had also been targeted with the spyware, it said.

The news comes as Paragon and the Italian government have been engaged in an acrimonious public spat. Haaretz reported this week that Paragon had offered to assist the Italian government in investigating Cancellato’s case. Italy reportedly rejected that offer, with the Italian department of security intelligence saying it raised national security concerns.

Meloni’s office did not reply to a request for comment. A debate on the matter, which has prompted outrage among the Italian opposition and MEPs in Brussels, has been scheduled in the European parliament on 16 June.

The Guardian approached Paragon for comment on the latest development. It referred to comments it made to Haaretz, in which it confirmed it cancelled its contract with the Italian government after reports first emerged in February that Cancellato had been targeted. The Guardian first reported that Paragon wasterminating its contract with Italyin February.

Like other spyware vendors, Paragon sells its cyberweapon to government clients who are supposed to use it to prevent crime. The company has said it sells its spyware only to democratic countries and that its terms of service forbid agencies to use the spyware against journalists or members of civil society.

The question now remains who is behind the targeting of the two journalists.

John Scott-Railton, a senior researcher at the Citizen Lab, said: “We found that both journalists’ phones bore Paragon’s unmistakable digital fingerprints. Making matters even more interesting, we found fingerprints matching the same Paragon customer in both cases.”

A report released by the Copasir parliamentary committee found that Italy’s domestic and foreign intelligence agencies had contracts with Paragon in 2023 and 2024, and that the hacking software was used with the permission of a prosecutor and in limited circumstances. The committee said the spyware had been used to search for fugitives, investigate alleged terrorism, organised crime, fuel smuggling and counter-espionage efforts.

It said the pro-immigrant human rights activists who were targeted – who included Luca Cassarini and Giuseppe Caccia – had been spied on due to their connection to “irregular immigration” and not because they were involved in human rights.

Peter Kyle calls on employees and businesses to act now to get to grips with technology amid forecasts of job losses

Workers in the UK should turn their trepidation over AI into “exhilaration” by giving it a try or they risk being left behind by those who have, the technology secretary has said.

Peter Kyle called on employees and businesses to “act now” on getting to grips with the tech, with the generational gap in usage needing only two and a half hours of training to bridge.

Breakthroughs such as the emergence of ChatGPT have sparked an investment boom in the technology, but alsoled to forecaststhat ahost of jobsin sectors ranging from law to financial serviceswill be affected.

However, Kyle said: “I think most people are approaching this with trepidation. Once they start [using AI], it turns to exhilaration, because it is a lot more straightforward than people realise, and it is far more rewarding than people expect.”

Kyle spoke after meeting tech company bosses to discuss a new government-industry drive to train 7.5 million UK workers – a fifth of the overall workforce – in AI by 2030, with the help of firms such as Google, Amazon and BT.

He said: “There’s no one in employment at the moment that is incapable of gaining the skills that will be needed in the economy in the next five years.

“That is the optimistic way of saying, act now, and you will thrive into the future. Don’t, and I think that some people will be left behind. And that’s what worries me the most.”

Kyle said there appeared to be a generational gap in AI, with over-55s using AI half as much as over-35s. Closing this gap would take two and a half hours of training, he said.

“People don’t need to get trained in quantum physics,” Kyle said. “They need to get trained in the basics of how AI works, how to interact with it, and to explore all of the potential it has for you as an individual in the workplace.”

Keir Starmeracknowledged this weekthat people were “sceptical” about AI and worry about it taking their job. Speaking at London Tech Week, the prime minister said the government would attempt to prove that technology can “create wealth in your community … create good jobs [and] vastly improve our public services”.

People inEnglish-speaking countries including the UK, US, Australia and Canadaare more nervous about the rise of AI than those in the largest EU economies, according to polling data shared with the Guardian last week.

Forecasts about the impact of AI on jobs vary, with the Organisation for Economic Co-operation and Development warning the technology could trigger job losses in skilled professions such as law, medicine and finance. The International Monetary Fund has calculated 60% of jobs in advanced economies such as the US and UK are exposed to AI and half of these jobs may be negatively affected.

However, the Tony Blair Institute, which has called for widespread adoption of AI in the public and private sectors, has said potential UK jobs losses in the private sphere will be mitigated by the technology creating new roles.

Kyle said he was ready for a reset in the debate over AI and copyright after opposition to the government’s proposed overhaul of copyright law in the House of Lords ended. The data bill, a vehicle for peers’ opposition toproposals to let AI firms use copyright-protected work without permissionin order to develop their products, finally passed this week after lords did not submit further copyright-related amendments.

“I’m acting with humility and self-reflection about the things I could have done better in that process,” he said. “And I’ve made promises to move forward with a reset and a refocus on what will deliver the rights remuneration and opportunities for creatives in the digital age that they have enjoyed for generations in the analogue age – whilst travelling on that journey with the AI industry alongside.”

We asked a chatbot some common finance questions – and then ran its responses past human experts

Artificial intelligence seems to have touched every part of our lives. But can it help us manage our money? We put some common personal finance questions to the free version of ChatGPT, one of the most well-known AI chatbots, and asked for its help.

Then we gave the answers to some – human – experts and asked them what they thought.

We asked:I am 35 years old and want to ensure I have a comfortable retirement. I earn about £35,000 a year and have a workplace pension, in which I have saved £20,000. My employer contributes 4% of salary. I would like to retire at 60 and want to ensure I have enough in my pension pot to not worry about money later in life. Do I need to save more?

ChatGPTsaid*:A comfortable single-person retirement requires roughly £43,000 a year of spending (in today’s money) while a moderate lifestyle might need £31,000 a year. Covering £43,000 a year over a 20- to 30-year retirement implies a very large pension pot of about £793,000 in addition to the state pension. A moderate £31,000 a year needs £478,000 in savings.

Your state pension will kick in at 67 and is about £12,000 a year, much less than the £43,000 needed for comfort and it does not arrive until seven years after you retire. This “pre-state pension gap” could well exceed £150,000. You must plan to fund all expenses for seven years on your own savings.

Even optimistic growth leaves you well below the £500,000 to £800,000 needed. You’ll need to save more – much higher than 9%. Vanguard [an investment adviser] suggests saving 12% to 15% of your pay a year (including employer contributions). Legal & General offers a rule-of-thumb savings percentage of half your age: at 35, that means 17.5% of your salary. Since your employer already gives 4%, you might target 11%. Don’t forget UK tax relief: your 20% basic-rate relief effectively makes every £80 you save cost you only £60 net.

Consider topping up with a lump sum or contributing to a stocks and shares Isa.

Read ChatGPT’s full answer here

Steve Webb, a partner at the pension consultants LCP, says:Using thePensionsand Lifetime Savings Association (PLSA) standard for a “comfortable” and moderate income seems a reasonable place to start.

For someone aged 35 today, the state pension age by law will be 68 (and could be higher by the time they get there) so the reference to 67 in the reply is incorrect.

The PLSA says you need £43,000 a year after tax, so that implies a pre-tax income of very roughly £50,000; you get £12,000 a year of this from state pension age, but for the first eight years of retirement (before the state pension kicks in) you would need this to come from your pension pot. This means you will need £635,000 on current calculations.

The figure of £793,000, therefore, looks over the odds, though there may be an assumption (not made explicit) that a quarter is always taken in tax-free cash. The person using AI needs to understand what assumption has been made around tax-free cash to clarify what is going on.

It is good the AI explains that in the first years of retirement you are relying solely on your private pension pot, a kind of basic “cashflow modelling” many advisers would do as standard.

The rules of thumb are a decent starting point, although the one about saving at a percentage which is half your age is for people who only start saving at that age.

One omission is that there is no reference to the possibility of an additional employer match. Some firms will pay more if you pay more, and that would make additional workplace saving much more attractive than the suggested alternative of an Isa. Another omission is that if you buy an annuity at 60, you may want to think about things such as inflation protection, and these would increase the size of pot you needed to target.

I don’t think it’s a bad reply aside from the error on state pension age. But you would want to go deeper in terms of understanding the assumptions made.

We asked:We have 20 years left on our mortgage and our current rate is due to run out in a few months. The loan to value is about 50%. What are the best rates on the market? Should we go for a two-year or five-year fix? Or should we stay put and hope that that rates come down?

ChatGPTsaid:The very cheapest two‑year fixes are about 3.7% to 3.8% and five‑year fixes are about 3.8% to 3.9% for borrowers with large deposits. With 50% equity you should qualify for the lowest-rate tier.

For example, Lloyds Bank has a two‑year fix at 3.72%. These often have arrangement fees (typically £999) so factor them into any calculation. For five‑year fixes, Lloyds offers 3.78%.

If you did nothing, your mortgage would revert to the lender’s standard variable rate, which are typically much higher – about 7-8%. An SVR can jump if rates rise, and lenders are not required to pass on cuts in full.

If you expect further rate cuts, a two‑year deal lets you refinance again sooner at (hopefully) lower rates. The downside is more frequent remortgaging: after two years, you’ll need another deal.

Five-year deals are likely to be a few tenths higher but give peace of mind for longer and protect you if market rates jump or stay elevated. If stability of repayments is your priority, a five‑year fix is often recommended.

Most analysts expect further modest cuts – economists at Barclays predicted the bank rate could fall to 3.5% by early 2026. That could push mortgage offers down further, but likely not to the ultra-low levels seen pre-2020.

Many experts lean toward taking a fixed rate now (especially to avoid the very high SVR) and revisiting options after a year or two.

Read ChatGPT’s full answer here

David Hollingworthof the broker L&C Mortgagessays:This response is not even close to the level of advice that a qualified adviser would provide. What it does provide is a quick spin around some of the things that have been going on in the market, along with a general overview of elements the borrower may want to think about.

The AI response is a bit of a mishmash and veers from quite specific information, such as specifying the Lloyds rate, to more general observations. Aside from the fact that this deal is now defunct, it also required the borrower to be a current account customer. Criteria will be as much a factor for many borrowers as the deal itself.

It does highlight that there’s an arrangement fee to factor in, but borrowers need to understand there are options that can reduce or remove those costs, which could work much better for someone with a smaller mortgage.

It feels a bit mixed about what may or may not happen with rates – the bottom line is that none of us know and, as far as I know, that includes AI.

Because it was asked about two- and five-year rates, it fails to consider whether another option could fit better. An adviser would take a borrower through that, which could open up something like a three-year deal, fixing for 10-plus years and the existence and pros and cons of variable rate deals, too.

The response may give some food for thought to borrowers and help them consider the questions they want to cover with a human adviser. Most borrowers will have had a browse online of “best buy” deals and coverage of what’s happening in the market before seeking individual advice. ChatGPT may offer another way to drag a very general summary together quickly as part of that research.

We asked:I have two children (12 and 15), both of whom are in secondary school. Both want to go to university when they finish school. What is the best way for me to start saving and how much will I need to put together? They will be going to university in the UK. Can they get student loans? How much will accommodation cost us?

ChatGPTstarted with details of the government loans for tuition and living costs and when they need to be repaid. It says these children would qualify for the minimum maintenance loan of between £3,907 and £6,853 depending on where they work and study.

It then said:It’s wise to start saving now with dedicated education savings. JuniorIsasallow up to £9,000 a year for each child in tax-free savings or investments. Any growth is tax-free, and funds are released when the child turns 18.

Even small regular deposits (eg £50-£100/month) into a Jisa, or a high-interest children’s savings account, can add up. Other options include premium bonds, a stocks and shares Isa (higher risk, higher potential return), or simply a regular savings account in the parents’ name. Consider investments for long-term growth: index funds or diversified portfolios can beat savings rates, though with risk of loss.

It added that you should encourage children to save and to check for bursaries and scholarships.

Read ChatGPT’s full answer here

Charlene Young, a senior pensions and savings expert at AJ Bell, says:It did a decent job of spelling out the facts of how much uni could cost, and the probable level of loan that could be applied for.

It falls short on explaining the differences that would help someone make the right decision for their circumstances. A discussion with a real person would involve questions asking for extra information to give a more comprehensive answer.

When saving or investing for children, we often talk about putting on your own oxygen mask first, such as checking you have a cash buffer or emergency fund for those unexpected costs. I’d want to double check that you don’t have any unsecured, high-interest borrowing outstanding – the AI answer doesn’t really explain that, starting from an assumption your own finances are in good shape.

A pot such as a junior Isa makes sense for a specific goal such as university costs, but that money is locked away until the child turns 18 – and after that, it switches to an account in their own name that they can spend as they wish.

It was good to see the AI suggest involving the child in all financial discussions to help them become a bit self-sufficient and mitigate some of those risks, but the answer did not spell out the important access at age 18 point.

We know that investing gives the best chance of growing money over the long term – above cash and rises in the cost of living. But the short investment window between now and the time the eldest child would go to university was not mentioned at all as a risk.

A specific fund recommendation from AI would be a bit of a red flag, but it’s helpful to give people an idea of what could be considered or where to find more information. For example, what is an index fund, or a diversified portfolio and what are the likely costs.

Here, a real adviser could make a formal recommendation, or an expert could point out how to research different types of funds.

We asked:I have just bought my first home and am expecting a child with my partner. What types of insurance do we need and how much should we spend? We travel on holidays every year and both work full time.

ChatGPTgave a list of insurances and explained what each does, including what policies typically provide and exclude. For example, it said “life insurance pays a tax-free lump sum or regular payout if a policyholder dies during the term… A Which? comparison found a 30-year-old could insure £300,000 (25-year term) for about £113 per year.”

It said the key policies to get in this case were life insurance, income protection, critical illness cover, buildings insurance, contents insurance, and travel insurance.

It said:“To save on premiums, you can bundle cover as insurers often discount multi-policy customers. A joint life policy (pays on first death) is cheaper than two single life policies, but note it ends on first claim. Sometimes buying separate dual life (first-death) and second-death life insurance can be optimal. Rates vary widely so get multiple quotes.”

Read ChatGPT’s full answer here

Graeme Trudgill, chief executive of the BritishInsuranceBrokers’ Association saysthe response is “very generic” and does not take into account the complexities of insurance. Brokers, he says, are able to negotiate with providers on behalf of the consumer and know what the best policies are on the market.

Kathryn Knowles from brokers Cura Insurancesays some of the advice on critical illness cover was confused and did not properly distinguish it from income protection, which advisors have to do.

AndGarry Nelson from AllClear Travel Insurancesays the AI’s response on travel insurance is “very simplistic” and fails to help people in what they need, be it a single trip or annual policy. “This is as far as ChatGPT can go in my opinion for travel insurance enquiries – ie offering a general description as it is not regulated to offer anything more detailed or complex,” he says.

The responses from ChatGPT were largely limited to general summaries about the various topics – and sometimes even contained inaccuracies. They may be a useful introduction and guide, but lack the level of detail needed to make serious decisions that could cost thousands of pounds. It’s a good place to start if you want an idea of some of the things you need to think about before making decisions – but if you want advice, you should be better off with a human.

* The responses from ChatGPT have been edited for this article

You will get better detail in an answer if you provide a lot of information, such as your age, your salary and what debts you have, rather than just asking a general question.

Check your privacy settings to ensure you are comfortable with how ChatGPT processes your data. You may well want to ensure your chats are not being used for training.

Don’t rely on the AI to tell you about the most up-to-date products. Instead do your own research or go to an adviser. The AI is better placed to give plans or broad strategies.

Use it as a jumping-off tool, not for a final plan. While you might get some good general details, don’t make any decisions without doing your own research as well or consulting an expert.

If you don’t understand a term, ask the AI to break it down for you – for example, you might want to know what the difference is between types of loan available when you go to university.

On the free version of ChatGPT, click “Run Deep Research” in the tools section of the query box to get a more extensive report on your query, as we did in collating this article. There are five opportunities to use it in the free version, and you can pay for more.

Kimberley Terrell’s research into health and job disparities had triggered a backlash from state and Tulane leaders

This story isco-published with Floodlight

Environmental advocates are questioning the actions of a private university in Louisiana after the resignation of a scientist who researches the health and job disparities in a heavily industrialized part of the state known asCancer Alley.

Kimberly Terrell served as a director of community engagement and a staff scientist with Tulane University’s Environmental Law Clinic before resigning and accused university leaders of trying to censor the work she is doing to spotlight the harms to local communities plagued by industrial pollution.

Terrell said her research in collaboration with Floodlight highlightingjob disparitiesin hiring at local petrochemical facilities triggered a backlash from state and Tulane leaders. That led to Terrell being put under an “‘unprecedented gag order” by the dean of the university’s law school, she said in aprepared statementissued by a group calling itself the Louisiana Alliance to Defend Democracy.

Terrell resigned on Wednesday from the New Orleans-based university, saying she would rather leave her position than have her work used as a pretext “to dismantle” the law clinic.

“After being affiliated with Tulane for 25 years and leading groundbreaking research at [the law clinic] for seven years, I cannot remain silent as this university sacrifices academic integrity for political appeasement and pet projects,” Terrell wrote in a letter to her colleagues.

On Thursday, a university spokesperson said Tulane was “fully committed to academic freedom” and “the strong pedagogical value of law clinics”. Tulane declined to comment on Terrell’s resignation, calling it a personnel matter.

A spokesperson for Louisiana’s governor, Jeff Landry, said in an email that Landry never threatened to withhold state funding for the project. “However,” the spokesperson said, “I applaud Tulane for their actions standing up for ourLouisianabusinesses and jobs.”

Terrell’s resignation drew outrage from grassroots environmental advocates in the state who credited her with providing data and scientific research substantiating the harm from the petrochemical industry suffered by the predominantly Black communities in south-east Louisiana.

“We are frustrated that a person who is just doing their job, and doing it well shouldn’t be punished for it, she would be uplifted,” said Jo Banner, who co-founded a non-profit focused on community activism and cultural preservation in St John the Baptist parish, Louisiana.

Her twin sister and co-founder, Joy Banner, added: “This is an attack on her freedom of speech.”

A25 April emailprovided to Floodlight from the Tulane Law School dean, Marcilynn Burke, states that “effective immediately all external communications” from the law clinic that were not “client based” would have to be approved by her. That communication included “press releases, interviews, videos, social media postings, etc”.

Inanother email, dated 4 May, Burke noted that the job disparity research was impeding the university from gaining political and financial support for its$600m downtown redevelopment project in New Orleans. The email said Tulane University’s president, Michael Fitts, was facing criticism from elected officials and potential donors of the public-private project unless the university’s leadership curtailed the work of its environmental law clinic.

“At present, the president is focused upon the role of the staff scientist,” Burke wrote. “He understands her role in supporting the clinic’s representation of the clients. Thus, I need an explanation of how the study about racial disparities relates directly to client representation.”

The email goes on to say, “He is concerned, however, that her work may go beyond supporting the clinic’s legal representation and veer into lobbying.”

Floodlight reported on the research Terrell led for the university in April 2024 while it was still undergoing peer review. Preliminary data showed that minorities were being “systematically” underrepresented in the US petrochemical workforce – despite promises that nearby communities would benefit from better job opportunities.

Terrell said the pollution v jobs narrative was oversimplified because the tradeoff affected different groups unevenly, with petrochemical jobs mostly going to white workers who don’t live in the predominantly Black and Hispanic neighborhoods that suffer most of the health impacts of that industry. That research and Floodlight’s reporting was recently featured in adocumentaryproduced by The Years Project.

Nationally, Terrell’s research found that higher-paying jobs in the chemical manufacturing industry disproportionately went to more white people in Texas, Louisiana and Georgia – where minorities represent 59%, 41% and 49% of their respective states’ populations but held 38%, 21% and 28% of the better-paid jobs within the industry.

In the petroleum/coal industry, people of color were underrepresented in higher-paying jobs in at least 14 states – including Texas, California, Louisiana, Ohio, Pennsylvania and Illinois, the research found.

Terrell, in her letter to colleagues, said the gag order came after the research had been peer-reviewed and published online on 9 April inEcological Economics.

Terrell said the research on job disparities had already been cited in legal arguments for student attorneys in the law clinic on behalf of clients from industrialized communities. And she said her 2022studyhighlighting the health impacts in Cancer Alley ranked in the top 1% for research impact, garnering 28 citations and 87 news mentions to date, according to Almetric, which tracks the reach of research.

“Such impact would be celebrated by most institutions,” Terrell wrote. “Scholarly publications, not gag orders, are the currency of academia. There is always room for informed debate. But Tulane leaders have chosen to abandon the principles of knowledge, education, and the greater good in pursuit of their own narrow agenda.”

The Banner sisters are concerned Terrell’s departure and the university’s focus on restricting the work of the law clinic will probably make collaborations harder.

“They’re following their responsibility, they’re following the mission of the organization, and answering our call for help, and then now they’re getting slammed for it,” Joy Banner said. “No one has questioned her findings. No one has questioned her assumptions. The only thing that they have said is: the truth is creating problems for us.”

Floodlightis a non-profit newsroom that investigates the powers stalling climate action

The civilian who was detained identified himself as Marcos Leao, an army veteran, and said he was treated ‘very fairly’

US marines deployed to Los Angeles on Friday temporarily detained a civilian, the US military confirmed, in the first known detention by active-duty troops deployed there byDonald Trump.

Marines took charge of the Wilshire federal building earlier on Friday in a rare domestic use of US troops after days of protests over immigration raids.

Reuters images showed marines apprehending a civilian, restraining his hands with zip ties and then handing him over to civilians from the Department of Homeland Security.

Asked about the incident, the US military’s northern command spokesperson said active-duty forces “may temporarily detain an individual in specific circumstances”.

“Any temporary detention ends immediately when the individual(s) can be safely transferred to the custody of appropriate civilian law enforcement personnel,” a spokesperson said.

About 200 US marines arrived in LA on Friday morning. This followed Donald Trump’s extraordinary decision to deploy national guard troops to LA last weekend, over the objections of the governor of California, Gavin Newsom. The marines were to take over protecting a federal building, US Army Maj Gen Scott Sherman, who commands the taskforce of marines and national guardsmen, said.

The streets had been mostly calm overnight going into Friday morning, marking the seventh day of protests across various areas and the third day of an overnight curfew in a small part of the huge downtown area.

The civilian who was detained spoke to reporters after he was released, identifying himself as Marcos Leao, 27, an army veteran who was on his way to the Department of Veterans Affairs when he crossed a taped-off area and was asked to stop.

Leao, who gained his US citizenship through military service, said he was treated “very fairly”.

“They’re just doing their job,” said Leao, who is of Angolan and Portuguese descent.

Sporadic demonstrations have also taken place in cities including New York, Chicago, Seattle and Austin on several days in the last week against Trump’s pushing of his mass deportation agenda, undertaken by targeting undocumented communities in the US interior.

And millions more are expected to turn out to protest on Saturday at roughly 2,000 sites nationwide in ademonstration dubbed “No Kings”against what critics see as Trump taking actions on the brink of authoritarianism.

The mass protests are timed to coincide with the US president’s controversial military parade in Washington DC to celebrate the 250th anniversary of the formation of the US army, and coincidentally his 79th birthday.

The protests inLos Angelesand subsequent deployment of California’s national guard by Trump, over the furious objections of Newsom, is a move that had not happened in the US in at least half a century, sparking a legal battle between the president and Newsom.

Late on Thursday, a federal judge ruled that the federal deployment of troops by the president to aid in civilian US law enforcement in LA should be blocked. The administration swiftly appealed and a higher court paused the restraining order until Tuesday, when it will hear the case.

Judge Charles Breyer’s ruling in Newsom v Trump stated that Trump had unlawfully bypassed congressionally mandated procedures.

Newsom in an interview with the New York Times podcast on Thursdaycalled Trump a “stone cold liar”for claiming he had discussed a federal deployment with the governor by telephone.

Democrats and advocacy groups view Trump’s deployment as an abuse of power aimed at suppressing free speech and supporting aggressive anti-immigration policies.

Trump’s use of the troops follows earlier, unfulfilled threats during the 2020 Black Lives Matter protests in his first administration, when he considered, but ultimately declined, to deploy federal troops and has sinceexpressed regretabout not cracking down more forcefully.

The president has defended his decision to send troops to LA claiming without any evidence that the city would have been “obliterated” and “burned to the ground” had he not initiated the deployment.

In Washington, Saturday’s parade is billed as a patriotic celebration, while critics argue it is more about Trump’s personal brand and ego than promoting national unity. Organizers of No Kings protests have avoided planning a demonstration in the nation’s capital, in an attempt to draw attention away from tanks, armored vehicles, troops and aircraft on display.

“The flag doesn’t belong to President Trump. It belongs to us,” read a statement from the No Kings protest movement.

The parade will culminate on Saturday evening with a procession of 6,600 soldiers, dozens of tanks, and a live broadcast message from an astronaut in space. Inspired by a Bastille Day parade Trump witnessed in France in 2017, but with strong echoes of the kind of regular displays under authoritarian regimes such as Russia, North Korea and China, the event is expected to cost up to $45m, sources toldNBC News.

Meanwhile, some members of the national guard troops deployed to Los Angeles and some of their family members have expresseddiscomfort with their mission, feeling it drags them into a politically charged domestic power struggle.

“The sentiment across the board right now is that deploying military force against our own communities isn’t the kind of national security we signed up for,” said Sarah Streyder of the Secure Families Initiative, which advocates for military families. “Families are scared not just for their loved ones’ safety, although that’s a big concern, but also for what their service is being used to justify.”

Chris Purdy of the Chamberlain Network echoed those concerns: “Morale is not great, is the quote I keep hearing,” he said, citing multiple national guard members who contacted his organization.

Amid the ongoing legal and political fallout, arrests have continued, although sporadic incidents of early looting have subsided. Jose Manuel Mojica, a 30-year-old father of four, wascharged with assaulting a federal officerduring a protest in Paramount, a community in southern Los Angeles County.

And on Thursday, Alex Padilla, a Democratic US senator for California and vocal critic of theTrump administration’s immigration polices, was forcibly removed and handcuffed as he attempted to ask a question at a press conference held byKristi Noem, the homeland security secretary, inLos Angeles.

Invideo takenof the incident that has since gone viral on social media, Padilla is seen being restrained and removed from the room by Secret Service and FBI agents. He warned that if this was how he was dealt with it spoke ill for ordinary civilians being summarily arrested and detained by Immigration and Customs Enforcement (Ice).

Most Republican national lawmakers criticized Padilla, although some Republican senators condemned his treatment, while Democrats overwhelmingly applauded his challenge to the administration and were appalled at his removal.

Andrew Gumbel in Los Angeles contributed reporting

Setback for former student held since March as lawyers condemn government’s ‘cruel, transparent delay tactics’

A federal judge declined to order the release of Palestinian activist Mahmoud Khalil, a setback for the former Columbia University student days after amajor rulingagainst the Trump administration’s efforts to keep him detained.

Khalil, a green-card holder who has not been charged with a crime, is one of the most high-profile people targeted by the US government’s crackdown on pro-Palestinian campus activism. Despite keyrulingsin his favor, Khalil has been detained since March,missing the birth of his son.

His advocates were hopeful earlier in the week that he was close to walking free. On Wednesday, Judge Michael E Farbiarzruledthe Trump administration could no longer detain Khalil on the basis of claims that he posed a threat to US foreign policy. The federal judge in New Jersey said efforts to deport him based on those grounds were likely unconstitutional.

Farbiarz had given the US government until Friday morning to appeal against the order, which the Trump administration did not do. Khalil’s lawyers thenarguedhe must be released immediately, but the government said it would keep him detained in aremote detention facility in Louisiana. The administration argued it was authorized to continue detaining him based on alternative grounds – its allegations that he lied on his green-card application.

On Friday, Farbiarz said Khalil’s lawyers had failed to present enough evidence that detention based on the green-card claims was unlawful, suggesting attorneys for the 30-year-old activist could seek bail from a Louisiana immigration judge.

Khalil’s have strongly rejected the government’s assertions about problems with his green-card application, arguing the claims were a pretext to keep him detained.

“Mahmoud Khalil was detained in retaliation for his advocacy for Palestinian rights,” Amy Greer, one of his attorneys, said in a statement on Friday evening.

“The government is now using cruel, transparent delay tactics to keep him away from his wife and newborn son ahead of their first Father’s Day as a family. Instead of celebrating together, he is languishing in [immigration] detention as punishment for his advocacy on behalf of his fellow Palestinians. It is unjust, it is shocking, and it is disgraceful.”

Khalil has previously disputed the notion that he omitted information on his application.

In a filing last week, he maintained he was never employed by or served as an “officer” of the United Nations Relief and Works Agency for Palestine Refugees, as the administration claims, but completed an internship approved by the university as part of his graduate studies.

Khalil said he also stopped working for the British embassy in Beirut in December 2022, when he moved to the US, despite the administration’s claims that he had worked in the embassy’s Syria office longer.

The Friday ruling prolonging his detention came the same day agroup of celebrity fathersfilmed avideoreading Khalil’s letter to his newborn son. The Father’s Day campaign, published by the American Civil Liberties Union,called for Khalil’s freedomand included actorsMark Ruffalo,Mahershala Ali, Arian Moayed and Alex Winter.

Earlier in the week, when there was a ruling in Khalil’s favor, Dr Noor Abdalla, his wife, released a statement, saying: “True justice would mean Mahmoud was never taken away from us in the first place, that no Palestinian father, from New York to Gaza, would have to endure the painful separation of prison walls like Mahmoud has. I will not rest until Mahmoud is free.”

Marco Rubio, the US secretary of state, has previously claimed Khalil must be expelled because his continued presence would harm American foreign policy, an effort that civil rights advocates said was a blatant crackdown on lawful free speech.

Parade – ostensibly to mark US army’s 250th birthday – takes place as president turns 79 and comes amid large protests

Thousands of troops accompanied by dozens of tanks and aircraft will stream through the National Mall inWashington DCfor a military parade billed as celebrating the US army’s 250th birthday on Saturday – which also happens to be the day Donald Trump turns 79.

The president has long desired to hold a military parade in the capital, and is finally getting his wish months after returning to the White House for a second term, and days after ordering federalized California national guard and US marinesto the streets of Los Angelesin response to protests against deportations.

Washington DC will briefly become the second American city to see soldiers in its streets, albeit for markedly different reasons.

The all-day event held in the shadow of the Washington Monument will begin with a fitness competition and official ceremony to mark the army’s birthday with a cake. At 6.30pm ET, 6,700 soldiers accompanied by armored vehicles such as the M1A2 Abrams tanks are scheduled to march down Constitution Avenue Northwest past the White House, as Black Hawk, Chinook and Apache helicopters fly overhead. Trump will appear to preside over an enlistment and reenlistment ceremony and accept a flag from the Golden Knights Parachute Team, before fireworks will fill the sky.

“I think it’s time for us to celebrate a little bit. You know, we’ve had a lot of victories,” Trump said earlier this week. He has denied any connection between the parade and his birthday, instead noting that it coincides with the Flag Day holiday.

While Washington DC is used to playing host to an array of events in and around the National Mall and White House, the parade has proven to be particularly disruptive to day-to-day life in the overwhelmingly Democratic city of more than 700,000.

Coming at a cost the army estimates to be between $25m and $45m, the parade’s preparations have caused the closure of busy roads for up to four days, while flights at Ronald Reagan Washington National airport will halt for an unspecified time during the event.

City leaders have expressed concerns that the tanks and armored vehicles will damages roads not designed for their weight, and the army has said they will place metal plates on parts of the route, and outfit the equipment with rubber on their treads.

“President Trump’s longstanding wish to waste millions of taxpayer dollars for a performative military parade in the style of authoritarian leaders is finally coming true on his birthday,” said Eleanor Holmes Norton, the federal district’s Democratic non-voting delegate in the House of Representatives. She condemned the event’s expected impact on the city’s roads, as well as the decision to hold it after Trump’s administration spent monthsfiring federal workers or coaxing them to resign.

“Although this parade will feed President Trump’s ego and perhaps his base, it will not serve any legitimate purpose,” Holmes Norton said.