Connect with us

Tools & Platforms

XAI releases Grok 4 amid furor over antisemitic comments

Published

on


Generative AI vendors xAI and Perplexity released new models and products to challenge mainstream vendors.

Amid controversy surrounding its Grok AI chatbot making a series of antisemitic comments, xAI released Grok 4 on Wednesday night.

During a live stream on X, xAI’s founder and X owner, Elon Musk, said the model can perform at a postgraduate level in mathematics, chemistry and linguistics based on tests like AI benchmarking platform Humanity’s Last Exam.

“With respect to academic questions, Grok 4 is better than a PhD level in every subject, with no exception,” Musk said during the livestream.

He added that while the multimodal generative AI model has not yet discovered new technologies, it could do so later this year or by 2026.

“AI is advancing faster than any human,” Musk said.

Meanwhile, upstart AI search vendor Perplexity released an AI browser.

Examining Grok 4

The new model has reasoning and problem-solving capability and uses DeepSearch to access factual information from the web, including the X platform. DeepSearch is a tool for web-based analysis and helps with complex queries that require multiple steps.

Grok 4 can process text and image inputs and has a new voice called Eve. The model can also perform multiple tasks simultaneously and is agentic, meaning it can use one or numerous agents for functions. It has a 256k context window and comes in standard and Heavy versions. Standard costs $30 per month, and Heavy costs $300.

The standard version performs single-agentic tasks, while the Heavy version is multi-agentic.

The release of Grok 4 comes only a few months after Grok 3 was released earlier this year, and days after Grok produced a slew of antisemitic responses.

While Grok 4 shows the progress xAI is making in foundation models, the uproar over the model overshadowed the latest version’s technical capabilities, said Arun Chandrasekaran, an analyst with Gartner.

“They have solid research and technical capabilities,” Chandrasekaran said.

Also, the benchmarks that xAI cites seem accurate, but enterprises should not make their decisions about models based on benchmarks, said Bradley Shimmin, an analyst with Futurum Group.

“It is a very much a guidepost, at best,” Shimmin said. “It tells us that Grok 4 aligns with other frontier-scale models.”

He added that the Grok models have been in line with other frontier models for some time, but the update with Grok 4 shows that xAI has been trying to improve the model’s ability to exceed other models on Humanity’s Last Exam.

Safe and responsible AI

Despite the advancement, xAI needs to focus on responsible and safe AI, according to many tech observers

“They need to focus more on guardrails,” Chandrasekaran said. XAI should concentrate more on safety and ensure that the safety mechanisms are layered as part of the entire process of training and releasing a model, including considering prompt inputs.

“Particularly in the case of Grok, it’s more about the recency,” Chandrasekaran said. This is because Grok seems to be taking context from the content coming from the X social media platform, known for its sometimes virulent and uncensored arguments about politics and culture. “They need to have a better filtering way from the context because otherwise the model could be very easily baited and biased from the recency of the inputs that are coming from X.”

In response to the comments the Grok chatbot made about the holocaust and false statements about “white genocide” in South Africa, xAI blamed a programming error.

But for some, model’s offensive hallucinations go beyond an error made by a computer system.

“This is just the latest instance in which [Musk’s] work and reputation are bound up with antisemitism,” said Michael Bennett, associate vice chancellor for data science and artificial intelligence strategy at the University of Illinois Chicago. “For the industry, it’s just a clear indicator that there’s still a lot of work to be done to get these models to produce useful, unbiased and socially acceptable responses. For his enterprises, it’s a further datapoint suggesting that his antisemitism perhaps is not a one-off.”

Permissiveness in the industry

The model’s responses also signal an attitude of laxness in the AI industry that has cropped up over the last year, said Kashyap Kompella, CEO of RPA2AI Research.

“The Grok incident is a sharp reminder that unfettered AI is a bad idea,” Kompella said. “Grok’s shenanigans expose the challenges of letting out AI chatbots unsupervised. We are ignoring and underinvesting in AI governance and guardrails. If there is a silver lining, this incident should wake up the AI industry to take AI governance seriously.”

Taking AI governance is especially important because these tools and technologies have a wider reach beyond the bounds of the U.S. and traditions of free speech, Bennett said.

“For technologies that enable speech that reaches a broader audience … the norms that we ought to be targeting to get the technology to align with, must necessarily be broad as well.”

The lack of governance could also affect xAI’s ability to attract enterprise customers.

“Model safety and responsible AI is a critical evaluation factor for a lot of enterprises; it’s an area where xAI needs to make a lot of progress if they want to be a serious enterprise contender,” Chandrasekaran said.

Perplexity AI

Meanwhile, Perplexity made good on its promise to launch an AI-powered web browser.

On Wednesday, the AI-search vendor launched Comet, a browser that Perplexity said is built for today’s internet.

The new browser is available to Perplexity Max subscribers and other select users by invite-only access.

With Comet, Perplexity appears to be to following a trend that has been developing with major search engines such as Google, Edge, and Safari, in which the value proposition is no longer the link the user has to click on, Shimmin said. Instead, the model is producing an answer, while the link might still be present in terms of a footnote.

“These are all merging into one user experience,” Shimmin said.

He added that it’s not clear whether Perplexity will disrupt existing search engines or user experience, but the vendor differs from traditional search vendors because it does not try to protect the existing search browser model. While other search engines are trying to protect the existing search system because of ad revenue, so Perplexity took a slower approach in embedding AI into search and started embedding AI right away.

Comet can connect with enterprise applications, including Slack, and users can ask questions with voice and text.

Esther Shittu is an Informa TechTarget news writer and podcast host covering artificial intelligence software and systems.



Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Tools & Platforms

AI technology drives sharp rise in synthetic abuse material

Published

on


New data reveals over 1,200 AI-generated abuse videos have been discovered so far in 2025, a significant rise from just two during the same period last year.

AI is increasingly being used to produce highly realistic synthetic abuse videos, raising alarm among regulators and industry bodies.

According to new data published by the Internet Watch Foundation (IWF), 1,286 individual AI-generated abuse videos were identified during the first half of 2025, compared to just two in the same period last year.

Instead of remaining crude or glitch-filled, such material now appears so lifelike that under UK law, it must be treated like authentic recordings.

More than 1,000 of the videos fell into Category A, the most serious classification involving depictions of extreme harm. The number of webpages hosting this type of content has also risen sharply.

Derek Ray-Hill, interim chief executive of the IWF, expressed concern that longer-form synthetic abuse films are now inevitable unless binding safeguards around AI development are introduced.

Safeguarding minister Jess Phillips described the figures as ‘utterly horrific’ and confirmed two new laws are being introduced to address both those creating this material and those providing tools or guidance on how to do so.

IWF analysts say video quality has advanced significantly instead of remaining basic or easy to detect. What once involved clumsy manipulation is now alarmingly convincing, complicating efforts to monitor and remove such content.

The IWF encourages the public to report concerning material and share the exact web page where it is located.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!



Source link

Continue Reading

Tools & Platforms

Google Cloud Summit London 2025: Practical AI deployment

Published

on


The level to which firms are already using AI varies according to technological maturity, strategic focus, and on an industry by industry basis.

But what’s clear is that from the smallest to the largest businesses, the landscape is already shifting. We’ve spoken about AI agents on the podcast before – the promise of autonomous AI activity – but it’s only now that businesses are beginning to put more faith in these tools.



Source link

Continue Reading

Tools & Platforms

AI on the line: How AI is transforming vision inspection technologies

Published

on


In an era of tightening global regulations and rising consumer expectations, the F&B industry is increasingly turning to advanced vision inspection technologies. From spotting defects to ensuring compliance, these automated inspection tools are reshaping quality control, enhancing efficiency, reducing waste and boosting safety. FoodBev’s Siân Yates explores how cutting-edge technology is reshaping the industry, one perfectly inspected product at a time.

In the food and beverage industry, traditional quality inspection methods have always relied on human observation – an inherently inconsistent and flawed process. Automated vision inspection systems offer a transformative alternative. By detecting foreign objects, assessing product uniformity and ensuring that only items meeting strict quality criteria reach consumers, these systems significantly enhance operational efficiency and minimise errors.

“As the food industry moves towards more automation, applications are becoming increasingly complex, largely due to the variability in food products,” said Anthony Romeo, product manager at US-based vision solutions company Oxipital AI. This complexity stems from the need for automated systems to adapt to the wide range of textures, sizes and ingredients in food, making precise automation a key challenge.

Stephan Pottel, director of strategy at Zebra Technologies, highlighted the rising demand for intelligent automation: “There’s a growing need for machine vision and 3D solutions, powered by deep learning, to address more complex food and packaging use cases, along with vision-guided robotics for tasks like inspection, conveyor belt picking and sortation workflows”.

Key features of vision inspection

1. Defect detection

Vision inspection systems excel in identifying defects that may go unnoticed by human inspectors. These systems utilise high-resolution cameras and advanced algorithms to detect foreign objects, surface defects, and inconsistencies in size and shape. For example, in the fruit packing industry, vision systems can identify bruised or rotten fruit, ensuring only high-quality products are packaged and shipped.

2. Label verification

These technologies are increasingly used for label verification, ensuring compliance with regulatory standards. Systems can check for correct placement, legibility and adherence to labelling requirements, such as allergen information and expiration dates. Vision is usually deployed for label verification, rather than food surface defects, enhancing compliance and reducing the risk of costly recalls.

3. Product uniformity assessment

Maintaining product uniformity is crucial in the food and beverage sector. Vision inspection systems can assess visual aspects such as size, shape and colour. For instance, a snack manufacturer might use vision inspection to ensure that chips are uniformly shaped and coloured, meeting consumer expectations for quality and appearance.

4. Adaptive manufacturing

Advanced vision systems, particularly those incorporating AI and 3D technology, enable adaptive manufacturing processes. These systems can adjust production parameters in real time based on the visual data they collect. For example, in a bakery, vision systems can monitor the size and shape of pastries as they are produced, allowing adjustments to baking times or temperatures to ensure consistent quality.

Advancements in AI

Recent advancements in AI, automation and 3D technology have greatly enhanced machine vision systems, increasing accuracy and providing realistic visual sensing capabilities. 3D imaging technologies are being used to assess the shape and size of products, ensuring they meet packaging specifications. For instance, in the seafood industry, 3D scanners can evaluate the dimensions of fish fillets, ensuring they are cut to the correct size before packaging. This not only reduces waste but also ensures consistency in product offerings.

What is more, 3D profile sensors improve depth perception and refine quality control, making them indispensable tools in industrial automation. Oxipital AI’s Romeo highlighted the potential of these technologies: “Removing defects before they reach customers is a key first step where vision inspection technology plays a role, but there’s even more data to be leveraged”. By preventing defects from the outset, manufacturers can boost yield and reduce waste.

AI-powered vision inspection systems can also facilitate real-time monitoring of production lines, identifying potential issues before they escalate. This capability allows manufacturers to implement predictive maintenance, reducing downtime and improving overall efficiency.

Zebra Technologies

AI and food safety

Consumer safety remains a top priority in the food and beverage industry. AI plays a crucial role in monitoring and analysing processes in real time, helping manufacturers navigate the complexities of compliance with legal requirements and certification pressures from major retailers.

As Zebra Technologies’ Pottel explained: “AI is ideal for food and beverage products where classification, segmentation, and object and anomaly detection are essential. It is also enhancing asset and inventory visibility, which is crucial for predicting contamination risks and maintaining high safety standards throughout the supply chain.”

“Vision technologies can help check the presentation of food products…offering a quick, repeatable and reliable way to assess the visual aspects of food products like size, shape and colour,” added Neil Gruettner, market manager at Mettler-Toledo Product Inspection.

He continued: “Deployment of this type of AI provides context to support rule-based machine learning and improve human decision-making. It also gives inspection equipment the tools to extract and interpret as much data as possible out of a product, facilitating the evolution and refinement of production processes through the continuous exposure to vast datasets.”

AI-enhanced vision systems also guide robots in handling food products, particularly those that are delicate or irregularly shaped. “AI has proved to be a great method for tackling applications with a high frequency of naturally occurring organic variability, such as food,”Oxipital AI’s Romeo explained, adding that this adaptability ensures gentle and precise handling, particularly important when sorting fresh produce or packaging baked goods.

Fortress Technology uses AI to reduce contamination risks and identify defects. The company’s commercial manager, Jodie Curry, told FoodBev: “Streamlining processes reduces the risk of contamination and ensures consistent quality. Implementing automated technology and digital tools helps identify inefficiencies and boosts responsiveness.”

Fortress Technology

The role of combination inspection systems

The integration of multiple inspection technologies into single systems is another key trend in this space. These systems integrate various inspection technologies, such as X-ray, checkweighing and vision inspection, to provide a comprehensive assessment of food products. By combining these technologies, manufacturers can ensure higher quality control, better detection of defects and more efficient production lines. This trend allows for more accurate and reliable monitoring, helping to reduce waste, improve safety standards and enhance overall product quality.

For its part, Fortress offers combination systems that enable comprehensive and multi-layered inspection. The company is already leveraging its proprietary data software package, Contact 4.0, across its metal detection, X-ray and checkweighing technologies. Contact 4.0 allows processors to review and collect data, securely monitor and oversee the performance of multiple Fortress metal detectors, checkweighers or combination inspection machines connected to the same network.

Oxipital AI

Deep learning and quality control

Deep learning is revolutionising visual inspection by enabling machines to learn from data and recognise previously unseen variations of defect As Zebra Technologies’ Pottel explained: “Deep learning machine vision excels at complex visual inspections, especially where the range of anomalies, defects and spoilage can vary, as is often the case with food.

This technology is vital for automating inspections and ensuring quality. Deep learning optical character recognition (OCR) also improves packaging inspection by ensuring label quality, regulatory compliance and brand protection. It can verify label presence, confirm allergen accuracy and prevent mislabeling.

“The goal is to strengthen quality control by capturing an image and processing it against set quality control parameters,” Mettler-Toledo’s Gruettner pointed out.

Vision systems are increasingly deployed for label verification, ensuring compliance with legislative food labelling requirements. The Mettler-Toledo label inspection portfolio features Smart Camera systems (V11, V13, V15) for basic label inspections, including barcodes, alphanumeric text and label quality. For more advanced applications, the PC based V31 and V33 systems offer a larger field of view, faster throughput and enhanced inspection capabilities.

Oxipital AI uses 3D product scans and synthetic data generation to eliminate the need for hand-labelling images. “All training is done at Oxipital AI, enabling food and beverage customers to deploy AI without needing a team of experts,” said Romeo. “Our solutions are designed for immediate impact, requiring no coding, DIY or machine-learning expertise to implement and maintain.”

Mettler-Toledo

Real-world applications and future prospects

According to Zebra’s Global Manufacturing Vision Study, which surveyed leaders across various manufacturing sectors, including F&B, 66% of respondents plan to implement machine vision within the next five years, while 54% expect AI to drive growth by 2029.

These figures, coupled with the expanding market for vision inspection systems, suggest

that the majority of manufacturing leaders are prioritising the integration of these advanced technologies, seeing them as crucial tools for both immediate improvements and long-term growth.

This shift is partly driven by increasingly stringent government regulations, which demand more accurate labelling and packaging. Many companies are already successfully leveraging AI to enhance their operations, particularly in labelling processes.

Despite its clear advantages, the uptake of AI has been slow. The main barrier appears to be cost. While the initial integration can be expensive, AI has demonstrated significant long-term cost savings, making it a worthwhile investment over time.

Zebra’s studies have shown that the pressure to maintain quality while managing fewer resources is intensifying for manufacturers. As a result, cost remains a significant consideration when implementing AI solutions.

Fortress recommends consolidating AI systems into a single interface, which helps reduce costs in the long term. Curry told FoodBev: “The future of our food supply chain depends on advanced inspection systems that enhance food safety, reduce product waste and require minimal factory floor space”.

She continued: “Combination systems offer the benefit of space efficiency, as all sales, services, parts and technical support are handled by one provider. A single interface simplifies training, improves operational safety and drives cost savings through faster installation and reduced training time.”

As AI continues to evolve, its role in vision and inspection is set to expand. Advancements in machine learning, sensor technology and robotics will lead to even more sophisticated and efficient inspection systems, raising quality and safety standards for consumers worldwide.



Source link

Continue Reading

Trending