IT News Review by Control F5 Software: Netflix says it used AI in Argentinian TV series

Netflix says it used AI in Argentine TV series

Netflix said it has used generative artificial intelligence to produce visual effects that appear on screen for the first time in one of its original series, using a technology that has caused industry-wide concern in Hollywood.

Co-CEO Ted Sarandos described AI as "an incredible opportunity to help creators make movies and shows better, not just cheaper."

Sarandos gave as an example the Argentine sci-fi series "El Eternauta (The Eternauta)," where creators wanted to show a building collapsing in Buenos Aires - a visual effect that would have gone over the project's budget. The creative team collaborated with Eyeline Studios, a production innovation group within Netflix, to create the dramatic scene using AI.

ChatGPT can now "think" and "act" for you after a new update

OpenAI has announced a new feature for ChatGPT that allows the popular chatbot to perform actions on behalf of a user. It's part of a broad industry trend that aims to change the way people get things done on the internet: the tech giants hope that instead of switching between apps and manually searching the web, we'll one day be able to rely on AI agents to do everything for us.

ChatGPT's agent model, which starts rolling out immediately, is yet another sign that big tech companies are increasingly investing in digital assistants with advanced capabilities. It also accelerates competition between OpenAI and Google, which has similar ambitions with its Gemini assistant.

OpenAI said on Thursday that ChatGPT's new agent mode "thinks" and "acts" using its own virtual computer, allowing it to handle complex and action-oriented requests.

A surveillance vendor has been caught exploiting a new SS7 attack to track the location of phones

Security researchers say they have caught a Middle Eastern surveillance company using a new attack capable of tricking cell phone operators into disclosing a subscriber's location.

The attack is based on bypassing security measures put in place by operators to protect access to SS7 (Signaling System 7), a proprietary set of protocols used by phone operators globally to route subscribers' calls and text messages.

DuckDuckGo now lets you hide AI-generated images from search results

Privacy-focused browser DuckDuckDuckGo is rolling out a new setting that allows users to filter AI-generated images from search results. The company says it's introducing this option in response to user feedback that AI images can be confusing when searching for something specific.

You can access the new setting by searching for an image on DuckDuckGo and going to the Images tab. There you'll see a new drop-down menu called "AI images." Users can choose whether or not they want to see AI-generated content by selecting "show" or "hide".

You can also enable the "Hide AI-Generated Images" option from the search settings.

OpenAI introduces a ChatGPT agent to solve complex tasks

OpenAI has launched a powerful new AI agent for its ChatGPT platform, designed to handle more complex and dynamic tasks. The launch, announced Thursday, marks an important step in the Microsoft-backed startup's efforts to lead in the evolving AI landscape.

Advanced AI agents - considered a more capable evolution of virtual assistants - are gaining ground in the tech industry. Large companies like Microsoft, Salesforce and Oracle are investing heavily in the technology to increase productivity and streamline operations.

The next iPad Pro could have a new camera

In Mark Gurman's Power On newsletter for Bloomberg, he notes that the public beta of iOS 26 is expected next week, along with a possible tweak for the upcoming iPad Pro with the M5 chip.

The iPad Pro with M4 launched last year moved the front-facing camera to landscape mode, since many users use the tablet that way or with a laptop-style stand, but the new model could also include a vertically-oriented camera so that video calls look good no matter how you hold the device.

Spotify's 30-hour audiobook plans are too short for long books

Spotify has launched two new "Audiobooks Plus" add-on subscriptions that allow premium users to double their audiobook listening limit to 30 hours. These are available to individual Premium subscribers as well as users managing Family and Duo plans. Other Premium account members can now request 15 hours of audiobook access from their plan administrator.

However, the new plans are still too short for those who prefer to listen to longer books.

The 30-hour limit isn't enough for titles like George R. R. Martin's A Game of Thrones or Brandon Sanderson's The Way of Kings, for example. You can listen to two or three shorter novels, but if you want to listen to them again in the future, you'll have to consume those hours again.

Substack wants to reinvent the wheel

In recent years, the newsletter platform has tried to expand into micro-blogging, TikTok creators and full-fledged websites. Now, the company is moving toward something its leadership has long criticized: advertising and social networking.

The New York Times writes that Substack is increasingly investing in its "Notes" feature, which resembles X (formerly Twitter) or Meta's Threads. Substack raised $100 million in a recent funding round. In June, the company said it has no plans to become profitable anytime soon.

Slack says its AI can understand your company jargon

Slack is using artificial intelligence to help business users understand confusing company-specific jargon faster and focus on the most important tasks. The communication platform announced it will "soon" add a feature that explains workplace jargon, as well as an AI assistant for Slack canvases, which automates repetitive tasks such as summarizing conversations and taking notes.

The AI context feature for messages will instantly explain messages that contain unfamiliar acronyms or phrases when the user hovers over them. For any internal or company-specific terms, Slack says the feature will use vocabulary and conversation history from the user's workspace to clarify project names, internal tools, or team-specific abbreviations. The goal is to save time without manually searching or asking a coworker.

Ring reintroduces video sharing with police

Ring is again allowing police to request video from users. Axon, a law enforcement technology company and maker of Tasers, announced in April that it is partnering with Ring to allow customers to share "relevant footage with law enforcement to help solve crimes faster and protect neighborhoods," according to a report by Business Insider.

The decision reverses Ring's previous plan to stop sharing video with police. Last year, the company discontinued the "Request for Assistance" feature, which allowed police officers to request footage from users' cameras through the Ring Neighbors app. At the time, the company said it would only allow requests from police in "emergency situations," which meant law enforcement could obtain images without a warrant anyway, raising privacy concerns.

Google's AI can now make phone calls for you

Google now allows all US users to use AI to call local businesses. The feature, available in Search, allows you to use AI to get pricing or availability information without having to talk on the phone.

Google began testing the feature in January and it's still only available for certain types of businesses, such as pet barber shops, dry cleaners and car services. When you search for one of these services, such as a canine barber shop, Google will display a new "have AI check pricing" button below the business listing.

Think twice before giving AI access to your personal data

Artificial intelligence is becoming increasingly present in almost every aspect of life - from phones and apps to search engines and drive-through menus. The increasing integration of AI assistants into web browsers shows just how much the way we search for and consume information has changed in recent years.

But this change also brings with it a worrying trend: AI tools are increasingly demanding extensive access to your personal data under the guise that they "need it to work properly." This level of access is far from normal - and should not be treated as such.

Microsoft warns companies and government institutions of attack on server software

Microsoft has issued an alert about "active attacks" on server software used by government agencies and companies to share documents across organizations and recommended security updates that customers should apply immediately.

The FBI said Sunday that it is aware of the attacks and is working closely with its federal and private-sector partners, but did not provide further details.

In an alert issued Saturday, Microsoft said the vulnerabilities target only SharePoint servers used inside organizations. It also noted that SharePoint Online in Microsoft 365, which is hosted in the cloud, was not affected by the attacks.

EU publishes compliance guidelines for high-risk AI models under new regulation

The European Commission has published new guidelines to help companies developing AI models deemed to pose systemic risks comply with the European Union's Artificial Intelligence Act (AI Act).

These guidelines aim to reduce developers' and companies' fears about regulatory requirements while providing more clarity on compliance. Companies that fail to comply risk significant fines - from €7.5 million or 1.5% of annual revenues, up to €35 million or 7% of global turnover.

The AI Act, passed last year, will apply from August 2 to AI models deemed to pose systemic risks, including basic general-purpose models developed by companies such as Google, OpenAI, Meta, Anthropic and Mistral. These companies have until August 2, 2026 to fully comply with the legislation.

Perplexity is in talks to pre-install its Comet AI browser on smartphones

Perplexity AI, the fast-growing Nvidia-backed startup that aims to rival Google in AI search, is in active talks with smartphone makers to pre-install its new mobile browser, Comet, on the devices, CEO Aravind Srinivas told Reuters on Friday.

If these partnerships are successful, Perplexity could significantly expand its user base by taking advantage of what is known as browser stickiness - the tendency for users to stick with the default browser on their phone. This could lead to frequent use of Perplexity's AI-driven features.

OpenAI launches $50 million fund to support nonprofit and community organizations

OpenAI, creator of ChatGPT, is launching a $50 million fund to support nonprofit and community organizations, the artificial intelligence company announced Friday.

The fund is the first action following a report with recommendations from the San Francisco-based company's nonprofit board, a committee formed in April to guide OpenAI's philanthropic efforts.

OpenAI is working to restructure its corporate model, which it says is necessary to continue to attract the huge amounts of capital it needs to remain competitive in the "AI race," a move it is trying to balance with its original mission to develop AI for the public good.

Meta apologizes after machine translation mistakenly announces death of Indian political leader

Tech giant Meta has apologized and said it has fixed a machine translation issue that caused one of its social media platforms to wrongly announce the death of Indian politician Siddaramaiah.

The chief minister of the southwest Indian state of Karnataka posted on Instagram on Tuesday in the local Kannada language, saying he was paying tributes to the late Indian actress B. Saroja Devi. He shared a similar message on Facebook and X.

Meta's machine translation tool, however, mistranslated the Instagram post, suggesting that Siddaramaiah, who uses a single name, was the one who had "passed away".

Summary compiled using a monitoring feed provided by Control F5 Software.