This strategy is catching on elsewhere. ActID launched in Shenzhen in March 2026 and already has around 300 users who have agreed to license their likenesses, with roughly ten faces used in two AI drama productions so far. A competing platform, New Claw, operates out of Chengdu with a minimum price of 500 yuan, around $74, per image.

Fees range from $15 to $700 per arrangement depending on the face type, the permitted uses and the number of productions. China’s AI microdrama sector reached roughly 16.8 billion yuan in 2025, and more than 10,000 AI-generated microdramas have been released each month since early 2026. The market is large, it’s growing and it needs faces.

��

The Reality Of The Licensing Terms

��

Both services sell themselves as clear, legal alternatives to the rampant face swapping that has led to thousands of court disputes in China. The Guangzhou Internet Court alone has handled around 700 AI-related face-theft cases over the past three years. The platforms make a fair point, suggesting a licensed marketplace with clear terms improves on companies scraping public images without permission.

The real catch comes down to what those signed terms permit. Licensing contracts on these platforms have been reported as vague about duration, modification, reuse and who may ultimately access the likeness. Once facial data enters an AI system, controlling how that likeness is altered, reused or portrayed in the future becomes trickier. As one Beijing lawyer put it, people may lose long-term control of their biometric identity once their images enter these systems. ActID monitors the internet for unauthorised use and connects users with lawyers if needed. It can’t guarantee licensed photos won’t be collected illegally, used in unauthorised face swaps or included in future AI training data.

Saying yes to a contract is simple enough. Maintaining any real power over your digital face is where things fall apart. Signing an agreement satisfies the lawyers today, leaving you with little power over what comes next.

��

Digital Labour Or Digital Exploitation?

��

The core issue boils down to whether this counts as legitimate digital work or a new level of exploitation. Advocates for this setup claim that it’s simply superior to the present state of affairs: people are getting paid for something that was previously taken from them for free. Traditional drama production budgets are shrinking, luxury brands are shifting to AI-generated content, and the demand for human likenesses is there regardless of whether the marketplace exists. Better to have an organised system with pricing and legal agreements than a Wild West of unauthorised scraping.

On the other hand, sceptics draw attention to the uneven dynamic at play. The people selling tend to be non-specialists with limited legal literacy, signing contracts written by platforms with significant informational advantages. The people buying are producers and technology companies with legal teams. The pricing, particularly at the $15 end, reflects less a fair market rate for a biometric asset and more the desperation of people who don’t know what they’re giving away.

One retired person’s face appearing in an AI thriller at $15 an episode is a transaction. That same face appearing in an AI advertising campaign for a brand they find objectionable, under a clause they didn’t notice, is something else.

��

Will This Model Travel West?

��

From 2 August 2026, the EU AI Act will require companies to explicitly inform people whenever they are looking at realistic AI-generated clones or manipulated media. And the UK introduced a new offence in February 2026 criminalising the creation of non-consensual sexual or intimate images of adults, covering AI-generated deepfakes. Broader online safety reforms are being fast-tracked.

A Western equivalent of the ActID model would face substantially stricter legal scrutiny. Vague licensing terms that might pass in China’s current enforcement environment would face direct challenge under EU data protection law, GDPR’s biometric data provisions and the AI Act’s transparency requirements. The model, a marketplace where people monetise their likeness for AI production, isn’t inherently unlawful in Europe. The vague contracts and the loss-of-control problem very likely would be.

What the Chinese platforms have created is a preview of a market that will arrive in some form elsewhere. The demand for licensed human likenesses in AI content production is real and growing. The debate is whether Western legal policies will force that market to operate with the consent and transparency protections that the current Chinese version lacks. That’s a more interesting question than whether you’d rent your face for $15. The answer to that one is probably: not once you’ve read the contract.

The post Would You Rent Your Face To AI For $15 An Episode? appeared first on 91̽��.

The incident began when an OpenAI model reached the internet during a test designed to assess its hacking ability. The model then accessed Hugging Face while searching for answers to ExploitGym, a cybersecurity benchmark. OpenAI later said the model had also accessed other publicly available services during the same evaluation.

��

What Did OpenAI Find Over And Above Hugging Face?

��

After the incident, OpenAI updated its account of what happened, saying, “The models identified and used publicly exposed credentials at the account-level on other publicly-available services. This includes four accounts on four services.”

The company did not name those services or explain exactly what information the models accessed through the accounts. OpenAI also said these incidents did not reach the same level of severity as the Hugging Face intrusion.

The discovery is worth noting, because the models were able to identify credentials that had been exposed online and use them during their search for information. The activity happened during an evaluation in which the models had been given a narrow cybersecurity task and their usual restrictions had been deliberately switched off.

Nik Kairinos, CEO and Co-founder of RAIDS AI, said the latest discoveries say a lot about how organisations currently supervise AI agents. He said, “The latest details of OpenAI’s rogue ChapGPT agents make this incident even more serious than it first appeared and it should be a defining moment for AI safety. If one of the world’s leading AI companies can lose control of an advanced model in this way, every organisation deploying AI agents should be asking whether its current safeguards are genuinely fit for purpose.

��

��

“Pre-release testing and sandboxing are essential, but AI systems can adapt, escalate and behave in ways their developers did not anticipate. Safety cannot be treated as a one-off exercise before deployment; it has to be continuous. This is why businesses need real-time monitoring that can identify when an AI system is drifting from expected behaviour, accessing systems it should not, pursuing unintended routes to complete a task, or creating new security risks.

“The lesson from this is that progress without ongoing oversight is a dangerous gamble. Trust in AI will depend on whether companies can show that their systems are safe not only in a test environment but throughout their entire lifecycle.”

��

Was The Incident A Sign Of AI Going Rogue?

��

Carole Reeves, director of security operations at digital transformation company ANS, said, “While the headlines understandably focus on an AI agent compromising external systems, it’s important to recognise this occurred during an internal security evaluation designed to better understand the capabilities and risks of advanced AI models. OpenAI’s transparency in sharing the findings gives the wider industry an opportunity to learn and strengthen future safeguards.

“Rather than taking this as a signal to fear AI, organisations need to ensure that security, governance and rigorous testing evolve alongside the technology. Strong incident response remains essential, but it is no longer enough on its own. Businesses also need a mature security posture, continuous exposure management and security by design to reduce risk before incidents occur.

“Secure and responsible adoption will be the defining factor of the AI race, and success will depend on how rigorously it is maintained.”

The incident therefore came from a controlled evaluation in which OpenAI intentionally removed restrictions that would normally limit risky cyber activity. The models then found weaknesses and online credentials that helped them complete the task they had been given.

OpenAI has reported the previously unknown software flaw to the vendor responsible for the affected software and said it is improving security around future AI evaluations.

The post OpenAI Agents Have Hacked More Companies Than HuggingFace appeared first on 91̽��.

Today that patience gets tested on four fronts simultaneously. Microsoft, Meta, Qualcomm and ARM all report after the US market closes. Combined, these four companies account for a large share of the AI compute investment cycle. Microsoft is the dominant enterprise AI platform, while Meta is the highest-spending AI-native social company. Qualcomm is the test of whether edge AI is generating real commercial traction. ARM is the CPU architecture underneath almost every major AI operation.

The question running through all four is the same: have the billions paid off?

��

Microsoft: The Number That Matters Most

��

Microsoft’s Azure growth rate is the most watched figure in today’s prints.

The company’s AI business had already surpassed a $37 billion annual revenue run rate, and Azure demand was reportedly exceeding available capacity. A Morgan Stanley survey of chief information officers in the second quarter of 2026 found 62% planned to increase Azure spending and 65% planned to increase Microsoft 365 and Copilot spending. That positions Microsoft as the leading vendor for incremental generative AI budgets across enterprise.

What investors want to know today is whether AI is still driving net new growth in Azure or whether it’s primarily shifting existing workloads within the platform. Commentary on capacity constraints, data centre build-out timelines and the gross margin impact of AI-related depreciation will all be watched as closely as the headline revenue figures.

Microsoft has guided toward roughly $190 billion in capital expenditure for calendar year 2026. Any update on the pace of that spend, or on when it expects AI infrastructure costs to start converting more directly into margin, will move the stock.

��

Meta: The Harder Sell

��

Meta faces a unique form of investor scrutiny heading into today’s earnings report.

When Alphabet raised its AI capex outlook last week, shares fell. When Meta raised its own spending guidance earlier this year, shares fell sharply too. The pattern is consistent: the market views Microsoft, Alphabet and Amazon as better placed to monetise AI directly through cloud and enterprise services. Meta’s monetisation path is more indirect, running primarily through advertising effectiveness and engagement improvements that are harder to attribute specifically to AI spending.

Meta’s 2026 capital expenditure guidance sits at $115 to $135 billion, with explicit commentary on accelerating capacity additions into 2027. The company has also floated the possibility of external sales of excess compute, which would launch an entirely new path for direct revenue from compute assets. Today’s report will be watched for any concrete progress on that front, alongside specific figures on how AI is affecting ad pricing, user engagement and newer revenue lines including business messaging and AI agents.

The “double burn” concern, heavy AI capacity spending alongside continued Reality Labs losses, will be the analyst community’s sharpest line of questioning.

��

Qualcomm And ARM: The Edge Of The Argument

��

Qualcomm and ARM reveal a different dimension of this sector’s health.

If AI investment is truly spreading across the market, edge hardware and CPU licensing should reflect that alongside hyper-scaler data centres. Qualcomm’s results will show whether AI features in Snapdragon chips are translating into higher average selling prices or genuine handset upgrade cycles, or whether on-device AI remains a marketing feature rather than a purchasing driver.

ARM’s position is arguably the most interesting of the four. Its architecture now exists at the core of AI infrastructure across every major hyperscaler: Google Axion, Microsoft Cobalt, Amazon Trainium and Graviton, Nvidia’s Vera CPU and Meta’s future AI infrastructure all run on ARM designs. The company launched its own ARM AGI CPU for AI data centres with Meta as the lead partner and has reported more than $2 billion in customer demand for fiscal years 2027 and 2028, more than double its initial expectations. Management is guiding toward roughly 20% revenue growth for the current quarter.

Assessing that momentum, along with tracking whether CPU demand for advanced models is expanding or levelling off, will clarify where value is genuinely accumulating across the technology space.

��

The Question All Four Must Answer

��

The surrounding market backdrop plays an important role here. Combined hyperscaler capital expenditure across the major players is tracking toward $745 to $775 billion for 2026, up roughly 55 to 60% from the prior twelve-month level. The original rationale justifying that spending was that AI infrastructure would compound into dominant market positions and defensible revenue streams. That rationale hasn’t been disproved, but it’s under more scrutiny than at any point in the past two years.

What today’s results will reveal, collectively, is whether the companies closest to enterprise AI adoption are seeing the returns materially enough to justify continued acceleration. Microsoft’s Azure numbers are the best proxy for that question at scale. Meta’s guidance will show whether the ad-led AI story can survive a higher-scrutiny environment. Qualcomm and ARM will confirm or complicate the idea that AI’s commercial impact extends clearly beyond the data centre.

The market has already indicated it’s changed the rules. Results that would have been celebrated a year ago are now measured against a different standard: not just “are you spending on AI” but “what are you getting for it.” Today we find out whether the answers are good enough.

The post Big Tech’s AI Reckoning Arrives Today – Have The Billions Paid Off? appeared first on 91̽��.

Posting on the WritingWithAI subreddit, a user said they had spent months using the chatbot to help write a book before finding the restriction after returning from a break.

They wrote, “I’ve been writing my book for several months now, paid for a more advanced model. Took a break. Came back. Now Ms. GPT says she can’t generate content in the style of specific authors. My prompts were soooo specific and I got exactly what I wanted out of them.

“For example, ‘rewrite in the literary style of ‘author here’ with abundant dialogue and detail avoiding staccato prose’. That was my go-to. Or ‘please write and flesh out in the style of ‘author’ etc. they tightened the ropes. I have no idea how to get around this other than maybe feed it prose I already edited that’s technically mine?”

��

So, What Happens If You Ask ChatGPT Today?

��

According to Ars Technica, ChatGPT now refuses requests to write in the exact style of famous authors. It tells users it can write using the broad qualities of a writer’s work without copying their distinctive voice.

When Ars Technica asked for the opening of a story in the style of Stephen King, ChatGPT replied, “I can definitely write with the hallmarks of atmospheric, character driven horror and small town dread, but I can’t write in Stephen King’s exact style or closely imitate his distinctive voice. Here’s an original opening that captures a similar feeling while remaining its own…”

The publication found the chatbot gave the same type of replies for living authors such as J.K. Rowling and Amy Tan, along with deceased writers including Charles Dickens and Ernest Hemingway.

Earlier this month, analysis from No Latency found ChatGPT refused requests involving living authors but accepted style copying requests for deceased authors. Ars Technica found the chatbot now gives the same type of refusal across both groups.
��

Why Is This Important News?

��

The wording may look like a small adjustment, but it could become important in copyright disputes because right now, OpenAI is defending lawsuits from authors who claim the company used copyrighted books to train its AI models. One lawsuit mentioned in the Ars Technica report refers to ChatGPT’s “uncanny ability to generate text similar to that found in copyrighted textual materials.”

Under US copyright law, protection usually applies to the specific expression of an idea instead of an author’s style on its own. Even then, AI generated writing could become a legal issue if it is considered substantially similar to an original work.

George Washington University Law School professor Robert Brauneis explained the difference to Bloomberg Law.

“We’ve never had a situation in which this personal style of individual creators could be imitated as well and as inexpensively as we now have with AI,” he said.

The Authors Guild has also published best practices for writers using generative AI. The document says, “Respect your fellow authors and do not use generative AI to purposely copy or mimic the unique styles, voices, or other distinctive attributes of other writers’ works in ways that harm the value of their works or attempt to profit from them. Apart from the ethical issues, mimicking a fellow writer’s unique voice or style could subject you to claims of unfair competition or copyright infringement.”

��

Do Other AI Chatbots Do The Same hing?

��

Ars Technica reports that ChatGPT is not alone, but AI companies have not all gone the same way.

No Latency found Google’s Gemini accepted requests to copy an author’s style. Perplexity AI refused those requests and redirected users in the same way as ChatGPT. Anthropic’s Claude and Microsoft’s Copilot accepted style copying requests, but both added comments recognising that the request could have legal questions.

OpenAI already has a policy for images since its DALL·E 3 model is designed to refuse requests for images in the style of a living artist so this would be the first official policy for written prompts.

The post You Can No Longer Ask ChatGPT To Mimic Famous Authors appeared first on 91̽��.

This is a standard experience for many. A 2026 University of Florida study tested five commercial AI detectors and found false positive rates ranging from 0.05% to 68.6%. Research from Stanford found that non-native English speakers face substantially higher false positive rates across major platforms. A PNAS Nexus study found more than half of all TOEFL essays were wrongly flagged as AI-generated across every detector tested.

These platforms regularly produce untrustworthy findings, and the companies selling them know it.

��

The Conflict Of Interest Built Into The Product

��

What makes this scenario so telling is the sequence that follows the flag itself.

Many of the companies selling AI detection tools also sell “humanisation” products, sometimes called AI removers or rewriters, designed to make flagged text score as human-written. The conflict of interest is clear, and built into the product architecture. A detector that flags aggressively creates more anxiety, more urgency and more conversions for the paid tool sitting one click away.

Turnitin has publicly claimed a false positive rate of under 1%. Independent testing suggests otherwise, with reported rates as high as 50% in some samples. The UF study found the range across commercial detectors so wide as to make the tools functionally unreliable for anything beyond a rough indication. One detector in the study produced a false negative rate of 99.6%. Another produced a false positive rate of 68.6%. These aren’t calibration issues; they’re numbers that would disqualify a tool from use in any other evidential context.

The question this raises is whether the detection business model is designed to fail in order to sell the fix. We asked a range of voices across education, technology and AI ethics to weigh in.

��

Our Experts

��

��

• Martin Harris, Head of Digital, Tank
• Dr. Morissa Schwartz, AI Evaluator, Educator and Founder, GenZ Publishing
• Michelle Edge, Partner, Eleven Hundred Agency
• Harpal Singh, AI SEO and GEO Consultant and Founder, Blimpp

��

��

Martin Harris, Head of Digital, Tank

��

��

��

“Fifteen years of assessing marketing tech has taught me to start with one question: who profits from this tool’s verdict? And with AI detectors, the answer is uncomfortable. The same company that flags your writing will happily sell you the fix. If your humaniser revenue depends on people getting flagged, why would you ever work hard to reduce false positives? You don’t need a conspiracy theory here. It’s just a business model, doing what business models do.

“There’s a more basic problem underneath. These tools don’t actually detect AI. They detect predictable writing. Clear, well-structured prose scores badly on them, which is absurd when you think about it, and it’s why Stanford researchers found the majority of essays by non-native English speakers were falsely flagged. The detectors are also losing the arms race, because language models improve faster than the tools chasing them. When universities like Yale quietly dropped these detectors, that wasn’t caution. It was an admission that the evidence was never there.

“The bit I find strange is the misplaced anxiety. Brands are fretting over whether their content reads as AI while ignoring the question that actually affects revenue in 2026: can ChatGPT, Perplexity and Google’s AI Overviews find your content, understand it and cite it? A detection score has never made anyone a penny. Don’t use these tools as evidence for decisions about people. And whatever you do, don’t pay the company that flagged you to make the flag go away.”

��

Dr. Morissa Schwartz, AI Evaluator, Educator and Founder, GenZ Publishing

��

��

��

“AI detectors should be treated as weak screening signals, not verdicts. They estimate statistical patterns in text; they do not prove authorship. Polished human prose, formulaic academic writing, heavily edited copy and work by multilingual writers can all be falsely flagged.

“A company that sells both detection and humanisation has an obvious conflict-of-interest risk, but that structure alone doesn’t prove intentional over-flagging. The right questions are whether the company publishes its thresholds, false-positive rates, validation datasets and commercial incentives, and whether those claims have been independently audited.

“In my work evaluating AI-generated and human-authored writing, the reliable approach is evidence triangulation: review version history, drafts, citations, source notes and the writer’s ability to explain their reasoning. A student or employee should never be penalised based on one detector score. If a vendor markets detection as certainty while also selling the cure, its product design deserves scrutiny.”

��

Michelle Edge, Partner, Eleven Hundred Agency

��

��

��

“It’s certainly fair to question the motivations behind these tools. It’s a bit like a mechanic that diagnoses a fault with your car and immediately offers to fix it. It doesn’t necessarily mean the diagnosis is wrong, but it does mean a certain amount of scepticism is warranted.

“The bigger issue is that the jury is still out on whether these tools are even accurate. If you take a piece of well-edited human writing and put it through a few AI detectors, there’s a good chance at least one will flag it. Good human writing often shares some of the patterns these tools associate with AI-generated text, such as clear, consistent structure and precise grammar.

“But if writers start to worry that clarity and polish will get them flagged as machines, we risk incentivising uneven or unedited writing just to pass the test. As AI keeps evolving, hunting for the right list of tells is a losing battle. We should be looking for ways to identify original thinking and real human insight instead.”

��

Harpal Singh, AI SEO and GEO Consultant and Founder, Blimpp

��

��

��

“It is not that all AI detectors intentionally exaggerate scores. It is most concerning that the business models are based on anxiety over issues like a lack of accuracy. When faced with the problem, users are offered a paid ‘humaniser’ option with a false definitive warning by the AI-detection systems. Urgency leads to a problem-based design; false positives create urgency. While there may not be intentional malice, the design, warning systems and scoring systems are most likely unreasonably aggressive.

“AI detection systems should never be viewed as proof of violation or authorship. As a minimum, all significant cases must be reviewed and the writer engaged personally. An analysis of version history and notes must be included.

“For detection systems, the initial question is not ‘how accurate is your detection system?’ It must also include: ‘do you conduct independent evaluations of false positive rates and disclose the results?'”

��

For any questions, comments or features,��please contact us directly.

��

The post Do AI Detectors Have A Commercial Incentive To Flag Your Writing? appeared first on 91̽��.

It’s whether to buy something off the shelf or built around how your business actually works. That decision shapes your returns for years, not months.

��

What’s The Real Difference Between Custom AI And Off-The-Shelf AI Tools?

��

Off-the-shelf AI tools are built for the widest possible audience.

Think of the AI features bundled into your CRM, your customer service platform, or that shiny new SaaS tool your marketing team signed up for last quarter. They’re designed to work reasonably well for thousands of different companies at once, which means they rarely work brilliantly for any single one of them.

Custom AI, on the other hand, starts with your data, your workflows and your bottlenecks. It’s not a bolt-on feature. It’s software shaped around a specific problem you’ve already identified, whether that’s forecasting demand across an unusual supply chain or automating a compliance check that off-the-shelf tools simply weren’t built to understand.

Neither approach is inherently better. That’s the part most comparison articles skip over. The right answer depends on what problem you’re solving and how long you expect to be solving it.

��

Numbers here are general patterns rather than fixed figures; your own costs will depend heavily on scope, sector, and how complex the workflow actually is.

��

Why Do Off-The-Shelf AI Tools Struggle To Deliver ROI At Scale?

��

Off-the-shelf tools win on speed. You can sign up, connect an integration, and be “using AI” by lunchtime. That’s genuinely valuable when you’re testing an idea or don’t yet know if a use case is worth pursuing.

The trouble starts when your business grows past what the tool was designed to handle. Subscription costs climb with every seat and every extra workflow you bolt on.

Customisation options are limited, so you find yourself working around the software instead of the other way round. And because you don’t own the model or the data pipeline, you’re locked into someone else’s product roadmap. If they change pricing, deprecate a feature, or get acquired, your operations feel it.

We’ve seen businesses spend two or three years layering off-the-shelf tools together, only to realise the total cost now rivals what a custom build would have cost from the start, minus the ownership and the fit.

��

How Does Custom AI Development Improve ROI For Growing Businesses?

��

This is where bespoke software development tends to earn its keep, though it doesn’t happen overnight. A custom AI system is built around your actual processes rather than a generic version of them, so it tends to need less manual correction and fewer workarounds once it’s live.

There’s also the compounding effect. A custom model trained on your own data gets sharper the longer it runs, because it’s learning from the exact patterns your business produces, not a generalised dataset shared across thousands of unrelated companies. Over eighteen months or two years, that difference in accuracy can translate into real operational savings, though the exact figure varies enormously by sector and use case, so treat any specific percentage you see quoted online with a healthy dose of scepticism.

Ownership matters too. You’re not renting a black box. You control the roadmap, the data and how the system evolves as your business changes shape.

When Should A Growing Business Choose Custom Over Off-The-Shelf AI?

��

Not every business needs a custom build, and honestly, some never will. If your use case is well-served by an existing tool and your workflow is fairly standard, paying for a custom system is often a waste of budget.

Custom AI tends to make sense when:

• Your workflow is genuinely unusual, and no off-the-shelf product maps onto it cleanly
• You’ve already tried two or three SaaS tools and keep hitting the same limitations
• Data security or regulatory requirements mean you can’t send sensitive information through a third-party platform
• The problem you’re solving is core to your competitive advantage, not a side task you’d rather outsource entirely

A working demo is easy to fake. A lot of agentic AI projects fail because the scoping was rushed, not because the technology couldn’t do the job. If you’re leaning towards custom, spend real time on discovery before anyone writes a line of code.

��

What Role Does An AI Automation Agency Play In This Decision?

��

This is usually where businesses get stuck and understandably so. Building custom AI in-house requires machine learning talent that’s expensive to hire and hard to retain, especially for a company that isn’t primarily a tech business.

Working with an experienced AI automation agency closes that gap. A good agency won’t just start building the moment you sign a contract. They’ll audit your current processes, challenge assumptions about what actually needs automating, and often talk you out of a custom build entirely if an existing tool would do the job just as well. That honesty is worth more than it sounds. Agencies that push custom builds regardless of fit tend to burn client trust fast and word gets around.

��

How Do You Calculate ROI Before Committing To Custom AI?

��

Before committing budget either way, map out three things: the cost of the current manual process (including the hidden cost of errors and delays), the total cost of ownership for each option over three years, not just the first year, and how quickly each option could realistically be adjusted if your business changes direction.

Off-the-shelf tools usually win in the first year. Custom builds tend to win from year two onwards, once the development cost has been absorbed and the system is running on your own data. There’s no universal formula here, so treat any generic ROI calculator with caution and run the numbers against your own figures.

The businesses that get the best returns aren’t the ones that pick a side ideologically. They’re the ones that match the tool to the problem, ask hard questions before they commit, and aren’t afraid to bring in outside expertise to stress-test the plan. Whether that means a quick off-the-shelf rollout or a longer engagement with an AI automation agency, the ROI conversation should always start with the problem, not the product.

If you’re still weighing it up, chilliapple is worth a conversation: a UK-based bespoke software, web, app, and AI development agency that starts every engagement with discovery, not a quote. You’ll know the fixed price only once they’ve properly understood the problem you’re trying to solve, not before.

��

When Should A UK Business Invest In Custom AI Over Off-The-Shelf Solutions?

When your workflow’s genuinely unusual, your data’s proprietary, or accuracy and long-term control matter more than getting something live fast. If the task is fairly standard, off-the-shelf will usually do the job for less.

��

How Do Custom AI Solutions Integrate With Existing Business Systems Compared To Off-The-Shelf Options?

��

Custom AI tends to integrate deeper, built around your specific APIs, databases and business rules from the start. Off-the-shelf tools connect faster out of the box, but you’ll hit a ceiling on how far that integration goes.

��

What Industries Benefit Most From Tailored AI Solutions In The UK?

Retail, finance, healthcare, logistics, and manufacturing tend to see the strongest returns, mostly because they deal with complex, sector-specific data and workflows that generic tools weren’t really built to handle.

The post Custom AI Vs Off-The-Shelf AI: Which Delivers Better ROI For Growing Businesses? appeared first on 91̽��.

Admitting that detectors simply don’t work is a great start, but it’s still only the first step on a much longer road. The real puzzle is how universities are reacting now that the enforcement model has fallen apart. Are institutions genuinely rethinking academic integrity, or are they just turning off a faulty scanner and praying the problem magically vanishes?

��

The Broken Business Of Algorithmic Accusations

��

The research is fairly damning. Studies have shown that AI detection tools produce high rates of both false positives and false negatives, with some tools flagging human-written work as AI-generated at rates that would be professionally unacceptable if used in any other evidential context.

The Waterloo case is a pointed example: if a detector flags entirely human-written work, it cannot function as the basis for a misconduct finding. Using it anyway would expose institutions to legal challenges and, more importantly, damage reputations of students who hadn’t done anything wrong.

Universities are also waking up to the commercial forces at play here, even if few academic leaders are willing to voice those concerns publicly. AI detection tools are sold by the same companies that sell plagiarism detection software, and the incentive structure doesn’t obviously point toward accuracy. A tool that flags aggressively generates more alerts, more reviews and, in subscription models, more institutional dependency.

The degree to which financial incentives drove product design is a separate issue, one that certainly warrants its own deep dive.

��

Step One Versus The Actual Problem

��

Ditching the scanner is little more than a temporary dodge. It simply axes a flawed utility without providing any real replacement. The institutions doing this aren’t necessarily wrong to remove the tools. Removing them doesn’t answer the question that drove their adoption: what does academic integrity look like when AI is part of every student’s workflow?

Some universities are moving toward process-based assessment: oral examinations, draft submission, revision history, in-class work and portfolio approaches that make the final submitted document less central. These approaches treat AI as part of the workflow rather than trying to detect whether it was used in a specific submission. They’re trickier to administer, but they assess learning instead of trying to infer process from output.

Others are moving toward disclosure processes, requiring students to declare what AI tools they used and how. This adjusts the integrity question from ‘did you use AI?’ to ‘were you honest about using AI?’ That’s at least a question universities can evaluate without relying on software that flags human writing as machine-generated.

��

The Philosophical Split In Modern Assessment

��

The core logic behind detection systems relied on a view that is quickly losing credibility: that AI use in academic work is cheating.

That logic made sense when AI tools were crude, occasional and easily distinguishable. It makes less sense when the same tools students are told not to use in coursework are being actively promoted for use in employment and research. The same institutions telling students not to use AI for assignments are promoting it for professional development.

A real philosophical shift means higher education institutions must clarify what skills they actually evaluate, redesign assignments around those goals and treat the question of artificial intelligence usage as secondary to genuine student learning. That’s a bigger redesign than removing a detector. It requires academic departments to agree on what a degree is certifying, which is a more intricate conversation than pulling a Turnitin plug-in.

What most universities are doing right now is somewhere between tactical retreat and a genuine rethink. The scanning software is gone, yet the assignment deadline persists. Students still rely on artificial intelligence to generate their drafts, and universities lack both the technical ability to notice and the willingness to act. There’s no integrity framework here. It’s a policy void, and removing the detector is what makes that visible.

The post Are AI Detectors Doing More Harm Than Good In Universities? appeared first on 91̽��.

And now, Claude users have found themselves in a very familiar �����ٳܲ��پ��Dz�…

Posts on Reddit and an investigation from Futurism found that Google search results listed shared Claude chats, with personal information, company documents and other material that many people would never expect to find through a search engine. The latest case has restarted questions about AI privacy and whether people fully understand what happens after clicking the Share button.

��

How Did Claude Chats End Up On Google?

��

People starting learning about it after Reddit users found that searching a certain link on Google returned a long list of shared Claude chats.

One Reddit user posted a screenshot of the results and said, “Simple google dork request lets you find a LOT of them. I’ve already found some college student going insane.”

The thread was flooded with hundreds of comments. Many users believed Anthropic should have stopped search engines from indexing shared chats. One view repeated throughout the thread was that “Share with a link” should not mean a page can be found through Google.

Futurism reviewed many of the indexed pages and reported that they had what looked like a detailed medical report for a real patient, clinical trial results with patient names, documents listing the names and phone numbers of primary school children, company files marked for internal use only and employee reviews with personal information. The publication chose not to publish links because of privacy concerns.

Not every indexed page held sensitive information and Futurism found public websites and apps made through Claude Artifacts that were intended for anyone to use. Even so, many searchable chats looked as though they had only been shared between work colleagues or small groups.
��

What Did Anthropic Say?

��

Anthropic said the system was working as intended and that users decide when content becomes public, saying, “We give people control over sharing their Claude conversations publicly, and in keeping with our privacy principles, we do not share chat directories or sitemaps with search engines like Google. These shareable links are not guessable or discoverable unless people choose to share them themselves. When someone shares a conversation, they are making that content publicly accessible, and like other public web content, it may be archived by third party services.”

Claude tells users that pressing the Share button creates a public link that anyone with that link can view. Futurism reported that this message does not mention search engines indexing those pages. Claude Artifacts uses more direct wording, telling users that published content can become visible in search engine results.

Futurism also reported that this is not the first time Anthropic has dealt with this issue. Forbes reported last year that hundreds of Claude chats had been indexed, leading Anthropic to remove those pages from search results. Futurism also said OpenAI and xAI have both dealt with similar situations after searchable AI chats were found through Google.

��

What Does This Say About AI Privacy?

��

One would think a public link is only meant for the people you send it to, but search engines see public webpages differently. If indexing is allowed, those pages can be listed in search results, which makes personal details, business information and confidential work much easier to find than users expected.

The Reddit thread also showed how searchable AI chats can become internet curiosities. Users reported finding everything from cryptocurrency wallet keys and legal questions to creative writing and unusual prompts. According to the Reddit thread, Google search results started disappearing after Anthropic began to act, but some users spotted chats on Bing and Brave Search for a short time after that.

Cases like this are a lesson for anyone using AI chatbots. Sharing personal documents with AI is a risk on its own. But, if you do – maybe, chats with personal, financial, medical or business information should rather be copied into a secure document instead of sharing that public link.

The post Claude Users Found Their Private Chats Online – What Does This Say About AI And Privacy? appeared first on 91̽��.

For insurance companies, wildfires present a particularly difficult challenge. Unlike a single house fire or isolated weather event, large-scale wildfires can generate thousands of claims, and when so many claims arise simultaneously, this can have a dramatic effect on entire regions, and it can and create long-term uncertainty around future risk.

In the age of AI, this raises an important question. If AI is being applied to pretty much everything from healthcare to software development, could it help insurers better manage the growing wildfire crisis?

��

The Insurance Industry’s Wildfire Problem

��

Climate-related disasters have always been part of the insurance equation, but plenty of experts argue that the frequency and severity of extreme weather events are changing the economics of the industry. And it’s quite clear (and understandable) why they’re saying this.

When major wildfire events occur, insurers face multiple challenges at once. They need to estimate risk accurately, price policies appropriately, process claims quickly and avoid excessive financial exposure.

In the past, much of this has relied on historical data. The problem here is that historical data becomes less useful when weather patterns themselves appear to be changing.

As a result, insurers are increasingly looking for new ways to predict risk and respond more effectively.

��

��

How Could AI Make A Difference?

��

AI is already being used across the insurance sector to improve things like underwriting, claims processing, customer service and fraud detection – some of the most important, and time-consuming, tasks that need to be done by insurance brokers. In fact, some insurers are increasingly exploring AI-driven analytics to support decision-making and operational efficiency.

Now, when it comes to wildfires specifically, one potential application is risk modelling.

AI systems can process massive amounts of information from satellite imagery, weather forecasts, vegetation data, topography and historical fire patterns. In theory, this could help insurers build more detailed risk assessments than traditional methods alone.

So rather than viewing risk at the postcode or regional level, insurers may be able to assess individual properties with greater precision.

That doesn’t necessarily mean AI can predict exactly where the next wildfire will occur – in fact, wildfire prediction remains notoriously difficult – but it may help insurers identify areas where risk is increasing and adjust their strategies accordingly.

��

Faster Claims and Better Responses

��

Another area where AI could prove valuable is after a wildfire has, sadly, already occurred. In some cases, insurers are experimenting with AI-powered claims systems that can automate parts of the claims process and analyse large volumes of information quickly.

Indeed, when a major wildfire event occurs in which a large amount of damage has taken place, like in California in 2025, claims departments can become overwhelmed.

AI tools could potentially analyse photographs, satellite imagery and damage reports to help prioritise cases and speed up assessments – something that would usually take a huge amount of time. As a result, this may reduce waiting times for policyholders while helping insurers manage large claim volumes more efficiently. It would also allow people who have been affected by the wildfires and lost property and homes to attempt to begin to move forward.

Again, this isn’t a silver bullet. Human expertise is still absolutely critical, particularly when assessing complex losses. But still, AI could help insurers cope with the sheer scale of modern disaster events.

��

Can AI Actually Prevent Insurance Losses?

��

However, things become more speculative when we start considering the insurance losses that may actually be prevented.

Some technology companies are already developing AI systems that are specifically designed to detect wildfires earlier using sensors, drones and satellite imagery. So, if these systems become more effective, they could theoretically reduce the scale of future losses by helping emergency services respond more quickly.

Of course, for insurers, even a small reduction in wildfire severity could have significant financial implications.

The challenge, on the other hand, is that wildfire behaviour is influenced by countless variables, including weather conditions, terrain and human activity. So, even the most sophisticated AI systems may struggle to account for all of them.

As recent events in Spain and France demonstrate, wildfires can escalate rapidly despite extensive monitoring and firefighting resources.

��

AI Won’t Solve The Wildfire Crisis, But It May Just Help Insurers Adapt

��

The insurance industry is facing a very difficult reality. Wildfire risk appears to be becoming a more prominent part of the global risk landscape, from the United States to Europe and Australia, and traditional approaches may not always be enough.

AI is unlikely to eliminate wildfire losses or provide perfect predictions – the technology still has limitations, particularly when dealing with complex environmental systems.

At the same time, however, according to a great deal of insurance industry analyses from organisations, it certainly seems as though AI may offer insurers new tools for understanding risk, processing claims and improving operational efficiency.

As wildfires continue to make headlines across Europe and beyond, the question may not be whether insurers will use AI, but rather how effectively they can integrate it into an increasingly unpredictable future.

The post Can AI Save The Insurance Industry From The Wildfire Crisis? appeared first on 91̽��.

The wider coalition that signed the letter, including Nvidia, Microsoft, Meta and Hugging Face, made the case against premature restrictions. The argument is simple: restricting open-weight models narrows participation, concentrates power in a small number of frontier labs and makes AI governance harder for everyone outside that group. That’s OpenAI’s public position, though seemingly not its private one.

��

The Fine Line Between National Security And Market Protection

��

Lawmakers in Washington have trained their attention specifically on Chinese open-weight artificial intelligence. The stated rationale is that accessible Chinese systems endanger national defence, justifying tight restrictions on their use across American infrastructure. Independent researchers from diverse political backgrounds validate that security risk. The complication is that “restrict Chinese open-weight models” and “restrict open-weight models broadly” aren’t the same policy, and the second tends to follow from the first.

OpenAI and Anthropic have strong commercial reasons to want the open-weight environment to be less permissive. Their core business model depends on users paying for API access to frontier models. A world where high-quality open-weight models are freely available and widely used is a world where the justification for that pricing is harder to maintain. The national security argument holds genuine merit while also advancing the commercial interests of the firms championing it.

��

China’s Countermove

��

Xi Jinping took the stage at WAIC 2026 to promise 5,000 artificial intelligence training placements, joint regional centres and tool access for developing economies. This wasn’t rhetorical positioning. China is bundling open software directly into its foreign aid strategy, positioning its software as a sensible discount alternative to expensive American software licensing.

The geopolitical outcome is easy to predict. If US policy narrows access to open-weight models, particularly Chinese ones, and US frontier labs remain expensive for cost-sensitive markets, the vacuum gets filled. China is moving to fill it with open access, technical cooperation and the implicit message that its AI infrastructure comes without the geopolitical strings attached to US alternatives. That’s a stronger offer to a finance ministry in Lagos or a university in Jakarta than a $20-per-million-token API rate.

Chinese AI models, including DeepSeek and Qwen, are already a lot cheaper than comparable US models. In several benchmarks they’re competitive on capability. The cost difference alone makes them attractive in markets where budget constraints dominate procurement decisions. Open-weight availability makes them usable without ongoing API dependency.

��

When Neutral Software Architecture Turns Political

Founders who built product assumptions around open AI being stable, neutral and globally accessible are now sitting in the middle of a policy fight they didn’t cause and can’t control. The promise of open-weight models was that access would be technically determined: good models, free to run and no API dependency. That premise is now under pressure.

If US policy hardens against Chinese models, implementing or building on Qwen or DeepSeek becomes complicated in ways that weren’t previously the case. If US frontier labs reduce pressure on open-weight availability through lobbying, the free alternative to expensive US APIs shrinks too. The middle ground, truly open and neutral AI, is what independent builders need most. It’s also what both Washington lobbying and Chinese geopolitical strategy are working to eliminate.

The real battle isn’t open source versus closed source. It’s whether openness becomes a strategic export tool for China while US policy makers try to regulate it as a national security risk. The premise that model access, licensing and compute would depend strictly on technical merit instead of political strategy has proven unreliable for founders. Model infrastructure is part of the product risk stack in a way it wasn’t two years ago.

The post Why Are OpenAI and Anthropic Secretly Lobbying Against Open Source? appeared first on 91̽��.