AI ON TRIAL

Merriam-Webster publisher sues Perplexity over content use

Just days after securing $200 million in new funding at a $20 billion valuation, AI-powered answer engine Perplexity is facing another lawsuit in the hot seat again.

Britannica, the publisher of Merriam-Webster and Encyclopaedia Britannica, has filed a copyright and trademark infringement lawsuit against Perplexity, charging the company with illegally using Britannica’s “trusted and human-verified content” and its historic brands, often verbatim and without consent.

Filed on Sept. 10 in a New York state court, the lawsuit references Britannica's 250-year-old history and its reputation for “trusted, fact-checked, meticulously researched content,” which it claims Perplexity is violating." Britannica further emphasizes the high demand for its content, noting more than 1 billion sessions on its website in 2024.  

The lawsuit charges Perplexity with “free riding” on the company’s investments, by “cannibalizing traffic” to Britannica’s websites with AI-generated summaries of its content. Unlike traditional search engines, Perplexity provides answers without sending users to other webpages, impacting web publishers, like Britannica, who rely on those clicks to sell subscriptions and advertising.

Britannica Group CEO Jorge Cauz said the company will take all steps necessary to protect its data and intellectual capital. 

"Perplexity claims to be the 'world's first answer engine' but the answers they provide to consumers are often Britannica's answers," he said.

The Britannica suit is the latest in a slate of copyright infringement lawsuits against Perplexity.

In late August, Japan news outlets, Nikkei Inc. and the Asahi Shimbun Co., filed a lawsuit in a Tokyo District Court for alleged copyright infringement, seeking an injunction and nearly $15 million each in damages. 

Dow Jones and New York Post parent company News Corp. sued Perplexity last October for “massive scale" copying of copyrighted news content, analysis and opinion.” In December, News Corp. amended its complaint, demanding a jury trial. Last month, Perplexity lost its bid to dismiss or transfer that case.

Other companies have also brought suits against Perplexity, including a "trademark infringement” filed in August by Perplexity Solved Solutions, accusing it of using its brand name. Perplexity Solved Solutions had been in business since 2017, a full five years before Perplexity was founded.

Perplexity isn’t alone in facing challenges when it comes to the use of AI-generated content, as copyright and content-use lawsuits have become a recurring theme across the industry.

Several AI companies, including OpenAI, have faced lawsuits from publishers over the use of copyrighted content to train their models. 

Ziff Davis sued OpenAI for using copyrighted content without permission. The New York Times also sued OpenAI and Microsoft for using its published content to train their technologies.  

These cases are unlikely to be the last, as developers continue to create more models and businesses increasingly commercialize and widely deploy AI technology. 

Last week’s announcement that Anthropic agreed to pay a record-breaking $1.5 billion to settle a class-action lawsuit from authors over copyright infringement clearly paves the way for more cases.

TOGETHER WITH IBM

AI + Automation: Winning the IT Game

Behind the US Open’s fan experience is a modern hybrid cloud approach infused with AI and automation.

The infrastructure scales on demand, transforms massive volumes of data into timely insights, and proactively manages performance to help minimize risk. The result is a seamless experience for millions of fans.

For business leaders, the takeaway is clear: with the right mix of IBM solutions such as cloud, AI, and automation, IT can become a driver of resilience, efficiency, and differentiation—not just an enabler of operations.

Read the full case study here

AI ETHICS

AI, suicide & Musk: Altman’s uncomfortable Tucker Carlson interview

If you haven’t watched or listened to Sam Altman’s interview with Tucker Carlson, you’re truly missing out. The OpenAI CEO fielded questions on seveeral high-stakes topics, including whether AI could surpass human power, the moral obligations surrounding ChatGPT, and even if an employee’s suicide was actually murder. 

The discussion was undeniably uncomfortable. Altman said ChatGPT is trained to reflect “the collective experience, knowledge, learnings of humanity,” but noted that aligning the model requires judgment calls. “We do have to align it to behave one way or another,” he said, pointing to the company’s internal “model spec” as the framework guiding what the AI can and cannot do. 

When Carlson asked Altman if ChatGPT has an “autonomy of spirit” that some users may even “worship,” Altman dismissed it, saying nothing about AI feels divine or spiritual to him. 

He then discussed his evolving views on the potential power of AI technology. Initially concerned about its concentration in a few hands, he said he now sees it as a means for broader empowerment by those using it.

The discussion also turned to ChatGPT and suicide, where Altman acknowledged the complexity involved with protecting user freedom, privacy and vulnerable users. He said, “I can imagine a world where if the law in a country says terminally ill patients have an option, ChatGPT could present it — here’s the law, here’s what you can do, here’s why you might not want to.”

The interview also delved into controversy and high-stakes conflicts surrounding an employee suicide, as well as Altman’s relationship with Elon Musk.

The interview took a darker turn when Carlson asked about the former employee’s death, framing it as a murder rather than suicide. Altman said the question felt “weird” and said it sounded like an accusation, stressing he was not responsible but understood why the family might be concerned.

Altman’s relationship with Elon Musk was also on the table during the interview. When asked about Musk, Altman described him as initially instrumental in OpenAI's creation, but now characterized him as a competitive presence attempting to slow the company down.

Check out the full interview: 

TOGETHER WITH KORE.AI

Big Enterprises Require Better Planning

If your company is enterprise sized – we’re talking thousands of employees, not hundreds – then implementing AI is a whole different beast. Between multiple offices/locations, countless teams, and endless different tech stacks and tools, it can be a real mess… which is why Kore.ai has made you this guide to help.

This free download gives you simple 5 step approach to scale AI implementations across your entire organization, no matter how big. It includes solutions for the biggest barriers you’ll face, as well as step-by-step approaches to how to embed AI into your company at its core. Deep View readers can download it for free right here.

AUTOMOTIVE

Volkswagen bets $1.1B on AI future by 2030

New cars weren’t the only item on the Volkswagen Group’s agenda at the vast IAA Mobility show in Munich last week. The German automaker made headlines with a pledge to invest $1.1 billion on AI by 2030.

The sum seems sizable given the group has made clear its determination to reduce costs, with operating profit down 15% year over year in 2024. But it’s this need for greater efficiency that has prompted the decision to escalate the usage of AI, with the company citing its potential to deliver “cost avoidance opportunities” of $3.9 billion by 2035.

More than 1,200 AI applications are already in use, but this number will be increased by hundreds more across all the company’s brands. Volkswagen Group’s huge portfolio includes Audi, Porsche, Bentley and Europe’s third biggest seller so far this year, Skoda.

No area will be left untouched by AI. In partnership with France’s Dassault Systèmes, Volkswagen is building an AI-powered engineering environment to cut vehicle development times by 25%, from around 48 months to 36 months, with virtual testing and component simulations to the fore.

In manufacturing, enhanced use of AI will be shared across more than 40 sites connected by the Group’s Digital Production Platform, a “factory cloud.” This will help optimize the interaction of complex assembly processes, lowering energy usage and material costs.

AI will assist with beefed-up cybersecurity, too, and Volkswagen is investigating the potential of collaboration with industry partners on an AI model trained on shared data that could benefit all participants.

In announcing the package, board member Hauke Stars called for more political support in Europe to advance AI innovation, and left no one in doubt about the company’s commitment to it. “Our ambition: No process without AI,” said Stars.

LINKS

• Google Gemini tops Apple’s Top Free Apps chart

• Elon Musk’s latest target is mobile networks

• AI is helping farmers battle climate change 

• Google Research introduces VaultGemma LLM

• Malaysia reins in data center growth

• AI fuels false claims after Charlie Kirk's death

• New AI chip powered by light delivers massive energy savings

• Fundlify: Crowdfund and presell products on Shopify

• AgentSea: Open-source platform for building, deploying and sharing AI agents

• xpander.ai: A full stack development platform for AI engineers

• Trace: Automate your busy work by routing tasks to the right agent whether human or AI

• Macaron AI: Your personal agent designed to help you live better

• GraphBit: An enterprise-grade LLM framework for building, deploying and scaling AI agents 

• AI SEO Score: Check your brand’s recognition, crawlability, and speed.

• EY: AI and Machine Learning Engineer

• Conveyor: Principal Product Manager, AI 

• AWS: Sr Generative AI Strategist, Generative AI Innovation Center

• Luma: Software Engineer, AI Agents

GAMES

POLL RESULTS

What's your biggest concern about the current AI investment landscape?

• Infrastructure spending is growing faster than actual AI revenues (14%)

• AI companies are making promises they can't deliver on (25%)

• The market is rewarding risk-taking over proven value (25%)

• AI technology isn't advancing fast enough to justify the costs (11%)

• Something else (9%)

• I'm not concerned - this is normal for emerging technology (16%)