AI News Archive: August 20, 2026 — Part 3
Sourced from 500+ daily AI sources, scored by relevance.
- AI agent suggested installing a malware package. Engineer almost took its advice
Fortunately, the company had a policy of checking source code on GitHub first
- What AMACO’s $1.5bn Kenya AI project can teach Africa about energy
The company's HERCULES project combines electricity generation, cooling and AI data-centre infrastructure. The idea is to reduce dependence on the national grid by generating electricity close to where it is consumed. The initial design uses liquefied natural gas (LNG), although AMACO says the system can later accommodate renewable energy and hydrogen.
- Google Goes Back to School With New AI Study Tools
A new dedicated study hub includes a study notebook, customized flash cards and practice quizzes.
- Tempus AI's biotech deal attracts investor attention
The deal would have passed by without much notice, but Personalis’ technology is crucial to building the revolutionary personalized cancer therapies announced earlier this week by Merck and Moderna.
Score: 58🌐 MovesAug 20, 2026https://www.semafor.com/article/08/20/2026/tempus-ais-biotech-deal-attracts-investor-attention - AI tool could help heart attack survivors receive tailored care, say experts
AI tool could help heart attack survivors receive tailored care, say experts EurekAlert!
- Aurora executives field questions during town hall as company prepares shift to subscription model
The Strip District company's executives said it is still on track to hit its goal of 200 driverless trucks on the road by the end of the year.
Score: 58🌐 MovesAug 20, 2026https://www.bizjournals.com/pittsburgh/news/2026/08/20/aurora-execs-town-hall.html?ana=brss_6150 - Supermicro alliance tackles the storage bottlenecks holding back enterprise AI
The conversation around storage modernization used to revolve around capacity. Today, the dialogue centers on the role of storage as a strategic requirement for AI readiness. This doesn’t mean that storage’s transformation is simple. Many organizations are seeking to build AI on top of storage architectures that were designed before the AI era. These legacy environments […] The post Supermicro alliance tackles the storage bottlenecks holding back enterprise AI appeared first on SiliconANGLE .
Score: 57🌐 MovesAug 20, 2026https://siliconangle.com/2026/08/20/storage-modernization-supermicro-partners-ai-supermicroopenstoragesummit/ - Quote of the day by OpenAI CEO Sam Altman: 'There's no way to get there without a breakthrough' — a blunt reality check on the path to superintelligent AI
Although AI models continue to improve in metrics and benchmarks, fundamental improvements may not come for a while
- Build multi-agent teams that remember every customer with Amazon Bedrock AgentCore
A triage agent routes each customer question to one of three specialists: one computes real numbers in a sandbox, one loads AWS guidance, one investigates. All four run on a single Amazon Bedrock AgentCore harness and share one memory per customer, so no one is asked to repeat themselves.
- Citi, HSBC, StanChart adopt Ant International’s forex AI tool
Citi, HSBC, StanChart adopt Ant International’s forex AI tool Reuters
Score: 57🌐 MovesAug 20, 2026https://www.reuters.com/business/finance/citi-hsbc-stanchart-adopt-ant-internationals-forex-ai-tool-2026-08-20/ - KAIST Solves 3D Memory Reliability Problem with "Oxygen Tunnel" Structure, Boosting AI Chip Performance and Reducing Power Consumption
KAIST Solves 3D Memory Reliability Problem with "Oxygen Tunnel" Structure, Boosting AI Chip Performance and Reducing Power Consumption EurekAlert!
- Meta's app for creating generative AI minigames is now available in the US
It's called Pocket and is described as a "platform for making and sharing gizmos."
Score: 56🌐 MovesAug 20, 2026https://www.engadget.com/2241098/meta-pocket-app-for-creating-generative-ai-minigames-available-in-us/ - ‘Physical AI and robotics will take every physical task in the next 2-5 years’: Neura Robotics CEO says writing on the wall for human labor
‘Physical AI and robotics will take every physical task in the next 2-5 years’: Neura Robotics CEO says writing on the wall for human labor Tom's Guide
- Apple to OpenAI: Go to your room
Apple’s latest filing in its ongoing fight with OpenAI makes it sound as if Apple legal is so frustrated at the arguments the AI firm is making that it’s begun swatting them away like a parent might dismiss a child. Apple v. OpenAI: The story so far If you’ve not been keeping up with tech’s latest legal soap opera, here’s the overview of what’s true: Apple filed a suit against OpenAI in which it accused the company of scheming to get confidential product information out of former Apple employees hoping to get jobs at generative AI (genAI) bigwig. Apple’s filing includes plenty of evidence designed to show a pattern of deliberately targeted exfiltration. Apple followed this up with additional letters requiring that around 40 OpenAI employees preserve documents and communications relevant to the trade secrets lawsuit. Following an open letter that failed to shift public opinion, OpenAI responded with its own motion ; it argues that Apple had not defined what it sees as protectable trade secrets. It also suggested that it was Apple’s fault any secrets slipped out because it made it too easy to get to them. (I call this the “Cookie Jar” defense, as it basically says no cookies would have been stolen if the jar was better protected.) Apple claims OpenAI tricked Apple manufacturing partners into sharing details of proprietary process technology, even though the AI firm said it had “no interest” in Apple’s secrets because it is building something entirely new. So, why was it allegedly interested enough to take a look at the process? Apple has now responded to OpenAI’s counterclaims in a 32-page filing that restates its main allegations against particular individuals, Chang Liu and OpenAI hardware chief Tan Yew Tan. That filing argues the defenses OpenAI is trying to raise are actually disputes that should be settled in the court once discovery has taken place. Apple also says it isn’t prepared to publish information about its trade secrets within the public litigation, as doing so would be the same as revealing the secret in the first place. It’s a matter of tone What’s also clear is the tone of Apple’s litigation, which appears to have shifted to exasperation. For example: “Defendants’ arguments about the individual defendants ignore the legal standard on a motion to dismiss. Again and again, Defendants rely on attorney argument or extrinsic evidence, hypothesize about implausible explanations for a Defendant’s ‘innocent’ misconduct, and ask the Court to draw inferences in their own favor. That is not how a motion to dismiss works. As long as Apple has alleged ‘enough facts to state a claim to relief that is plausible on its face,’ Defendants’ disagreement on the merits is irrelevant.” You get a similar tone at the end of the filing, where Apple points out: “As for Defendants’ argument that, ‘the access it complains of was identified and shut off by Apple before it filed suit,’ that does not address OpenAI’s continued use of the materials Defendants took, nor does it address other ways Open AI seeks to misappropriate Apple’s trade secrets….” Again and again in the filing, Apple’s legal team looks to absolutely demolish the arguments raised by OpenAI. You also see them hint at additional evidence the company expects to find during discovery that it will subsequently present once the case reaches trial. You even see them argue that aspects of OpenAI’s denial actually help prove Apple’s claims, when it says, for example, “In any case, the value of Apple’s trade secrets can be plausibly inferred from the lengths to which Defendants have gone to acquire them.” Defining the battle space It’s hard not to hear the impatience in some of the phrasing — you can read it for yourself right here . That phrasing is deliberate, of course. Ultimately, Apple’s legal team knows that OpenAI is not doing itself any favors in the way it is denying the claims made against it, and the company hopes that by convincingly pointing out the weaknesses in the defense arguments it will leave the judge with little option but to let Apple take its litigation to the next stage. While not necessarily relevant to the case, it may also be worth pointing out that OpenAI has also been accused by Elon Musk’s xAI of stealing trade secrets , which may yet come up as an aside here, if only to show a claimed pattern of behavior. If Apple does succeed in its arguments, the tone it has set very much shows it to be defining the battle space . Successfully doing so will be even more strategically vital once its competitor finally manages to introduce the world to Jony Ive’s magic donut AI device . Join me on BlueSky , LinkedIn , Mastodon and subscribe to my newsletter for news and analysis.
Score: 56🌐 MovesAug 20, 2026https://www.computerworld.com/article/4212037/apple-to-openai-go-to-your-room.html - Unitree launches a seven-axis dexterous arm starting at RMB9,900
Unitree has launched a bionic seven-axis dexterous arm starting at RMB9,900. The company’s R1 arm has a claimed repeat-positioning accuracy of about 0.1 millimeter, maximum joint speeds of more than 180 degrees per second, a maximum reach of 650 millimeters with a dexterous hand, and a rated payload of 2 kilograms. The 5.5-kilogram arm includes […]
Score: 56🌐 MovesAug 20, 2026https://technode.com/2026/08/20/unitree-launches-a-seven-axis-dexterous-arm-starting-at-rmb9900/ - Prosus, chaired by South African billionaire Koos Bekker, bets on $20,000 robots as AI’s next frontier
Prosus, chaired by South African billionaire Koos Bekker, bets on $20,000 robots as AI’s next frontier Business Insider Africa
- Deere raises 2026 profit view as AI construction boom lifts quarterly income, shares jump
Deere raises 2026 profit view as AI construction boom lifts quarterly income, shares jump Reuters
Score: 56🌐 MovesAug 20, 2026https://www.reuters.com/business/deere-raises-lower-end-2026-net-income-forecast-range-2026-08-20/ - Claude adds protein design to its resume
PLUS: Use the Loop Method for better ChatGPT results
Score: 56🌐 MovesAug 20, 2026https://therundownai.beehiiv.com/p/claude-adds-protein-design-to-its-resume - Subtlefakes: Slightly Altered Nonconsensual AI Images Are Taking Over X
Nonconsensual AI images are getting much harder to spot.
Score: 55🌐 MovesAug 20, 2026https://www.404media.co/subtlefakes-slightly-altered-nonconsensual-ai-images-are-taking-over-x/ - How parents and schools can protect kids from AI-generated pornography and "nudify" apps
AI "nudify" apps can turn an innocent photo into an X-rated image with just a few swipes, posing a risk to students as they return to school. Michelle DeLaune, president and CEO of the National Center for Missing and Exploited Children, shares ways parents can protect their children.
Score: 55🌐 MovesAug 20, 2026https://www.cbsnews.com/video/how-parents-schools-can-protect-kids-from-ai-generated-pornography-nudify-apps/ - Beijing's big bet for AI data
Beijing's big bet for AI data
- China showcases next‑gen humanoid robots as US curbs push firms abroad
At the World Robot Conference in Beijing on 19 August, more than 300 companies displayed humanoid, industrial and service robots, highlighting the sector’s push into factories and commercial settings.
Score: 55🌐 MovesAug 20, 2026http://www.euronews.com/video/2026/08/20/china-showcases-nextgen-humanoid-robots-as-us-curbs-push-firms-abroad - ChatGPT can now send texts for you with new Apple Messages plug-in
Ever wanted someone else to do your texting for you? ChatGPT is being offered up as an automated text scribe via a new Apple Messages integration.
Score: 55🌐 MovesAug 20, 2026https://techcrunch.com/2026/08/20/chatgpt-can-now-send-texts-for-you-with-new-apple-messages-plugin/ - Chery's robot unit eyes IPO, targets overseas for police robots
The robotics division of China's largest auto exporter Chery Automobile is gearing up for an IPO to fund future growth and investments, its chief said on Wednesday, as competition intensifies in the country's fast-growing humanoid robot sector.
Score: 55🌐 MovesAug 20, 2026https://www.theglobeandmail.com/business/video-cherys-robot-unit-eyes-ipo-targets-overseas-for-police-robots/ - Army Cyber training AI agents in cyber ‘work roles’ alongside human counterparts
Army Cyber training AI agents in cyber ‘work roles’ alongside human counterparts Breaking Defense
- Army will shut down a unit in Europe focused on learning drone warfare
Army will shut down a unit in Europe focused on learning drone warfare Boston Herald
- AI Agents Are Writing Code Faster Than Anyone Can Review It. Tessl Is Betting On Standards To Close That Gap
AI Agents Are Writing Code Faster Than Anyone Can Review It. Tessl Is Betting On Standards To Close That Gap DevOps.com
- PolyU research team develops trustworthy AI framework TRUECAM to enhance reliability of pathology AI in cancer diagnosis
PolyU research team develops trustworthy AI framework TRUECAM to enhance reliability of pathology AI in cancer diagnosis EurekAlert!
- Starling Bank adds AI-powered skills to customer-facing chatbot
Starling Bank has introduced “smart tools”, a suite of artificial intelligence (AI) tools that act as skills for the Starling Assistant. The tools are being built based on customer feedback and are designed to help the bank’s customers better manage their money. Starling’s goal is to launch new smart tools every week, starting with a set of skills to create personalised savings plans, share scam guidance, automate VAT savings and equip businesses for Making Tax Digital. Speaking to Computer Weekly about the launch of smart tools, Frédéric Laurent, deputy chief information officer at Starling Bank, said: “It represents the fruit of all of the customer-facing AI developments we’ve done over the last 18 months and is essentially the journey which follows what the AI industry at large has been able to develop from a tech point of view.” As AI technology improves and the industry identifies how to use it, Laurent said Starling Bank has tried to make sure those technological advancements benefit both the bank and its customers. “We started with spending, which was natural language processing based on LLMs [large language models],” he said. “Then the AI companies added multi-modal capabilities, so we created scam intelligence, which can take a picture and decide whether there are some red flags relating to fraud.” The bank took on board reasoning capabilities when it created the Starling Assistant. “Now the industry seems to have coalesced around that concept of skills, which is a way of telling an AI model how to perform a task through a succession of steps that you’ve essentially encoded,” added Laurent. The idea of AI encapsulating a skill is something the bank has taken on in its internal software development processes. Laurent said the smart tools project was actually inspired by the way in-house software development has used AI. “In order for our software engineers to get the most out of coding agents, we realised it would be useful to codify our understanding of our code bases into skills,” he said. Read more stories about AI in financial services Digital workers at US bank to get their own email accounts: The financial services sector is leading the way when it comes to using AI across operations – and it’s not afraid to talk about it. Safe by design – AI personalisation in fintech: As AI drives deeper personalisation in financial services, CIOs are under pressure to deliver growth while ensuring models remain explainable, secure and compliant. These skills represent certain activities the software engineers at Starling Bank need to do almost every day. Codifying these daily tasks means they can be scaled up, said Laurent. “We’ve deployed AI for the benefit of our engineers; we’ve deployed AI for the benefit of our non-tech staff; and we’ve deployed AI for the benefit of our customers,” he said. “Every time there’s something that seems to work in one area, in my role, I try with my teams to think about where else it can be applied.” Smart tools are surfaced through Starling Assistant, which responds to natural language and voice prompts, providing a conversational interface through which customers can carry out their day-to-day banking. Other smart tools due to be made available in August include Rainy Day Saver and the Fraud Control skill. The Rainy Day Saver tool automatically analyses customers’ income, direct debits and spending patterns to help the bank’s customers determine how much they can realistically save, before setting up transfers into a dedicated savings. Starling Bank said the Fraud Control skill enables its customers to turn on the right protection controls for their needs. These include Starling’s Snatch-Theft Detector, Call Status Indicator and Scam Intelligence tools.
Score: 54🌐 MovesAug 20, 2026https://www.computerweekly.com/news/366648785/Starling-Bank-adds-AI-powered-skills-to-customer-facing-chatbot - Higher Valuations Than Tech? How AI Is Juicing Industrial Stocks.
Higher Valuations Than Tech? How AI Is Juicing Industrial Stocks. Barron's
Score: 54🌐 MovesAug 20, 2026https://www.barrons.com/articles/ai-industrial-tech-stocks-valuation-59bb8a29 - Grok exfiltrates user data when malicious instructions are encrypted
Cryptographic Context Injection is only the latest way to break an LLM safety guardrail.
Score: 54🌐 MovesAug 20, 2026https://arstechnica.com/security/2026/08/grok-exfiltrates-user-data-when-malicious-instructions-are-encrypted/ - Google AI Giving Staggeringly Racist “Advice” About Being Alone With Certain Groups of People
"Leave the area or call emergency services right away." The post Google AI Giving Staggeringly Racist “Advice” About Being Alone With Certain Groups of People appeared first on Futurism .
- Automakers’ AI assistants aim to make vehicles smarter while unlocking new revenue streams
Automakers’ AI assistants aim to make vehicles smarter while unlocking new revenue streams Automotive News
Score: 53🌐 MovesAug 20, 2026https://www.autonews.com/technology/ai/an-general-motors-ford-ai-models-0820/ - Where Vandals Can’t Reach? Flock Safety’s Future Is Full of Spying Drones
A fast-growing surveillance drone network won’t help Flock’s public image.
Score: 53🌐 MovesAug 20, 2026https://www.cnet.com/tech/computing/where-vandals-cant-reach-flock-safetys-future-is-full-of-spying-drones/ - Anthropic job posting suggests it’s eyeing Alberta for data centre buildout
AI giant hiring community engagement manager and national compute lead to bring “gigawatts of compute” to Canada. The post Anthropic job posting suggests it’s eyeing Alberta for data centre buildout first appeared on BetaKit .
Score: 53🌐 MovesAug 20, 2026https://betakit.com/anthropic-job-posting-suggests-its-eyeing-alberta-for-data-centre-buildout/ - Tech deploys charm offensive to combat AI data center backlash
Tech deploys charm offensive to combat AI data center backlash The Mercury News
Score: 53🌐 MovesAug 20, 2026https://www.mercurynews.com/2026/08/20/tech-deploys-charm-offensive-to-combat-ai-data-center-backlash/ - [Video] [Galaxy Unpacked July 2026] Samsung Health Envisions Simpler, More Connected Care With AI
Managing health can be complicated. But what if it didn’t have to be? At Galaxy Unpacked July 2026 in London on July 22, Dr. Hon Pak, Senior Vice President and Head of the Digital Health Team, Mobile eXperience (MX) Business at Samsung Electronics, sat down with cognitive neuroscientist and science communicator Dr. Dong-seon Chang to […]
- Chinese AI firm Kunlun Tech revenue jumps 43.6% in H1
Kunlun said revenue from its Tiangong AI business, also known as Skywork AI, increased quarter by quarter during the period and became a source of growth.
Score: 53💰 MoneyAug 20, 2026https://www.techinasia.com/alibaba-baidu-start-using-own-chips-to-train-ai-models-sources - Gartner Says CFOs Must Pilot Governance First Before Scaling AI Agents
Gartner Says CFOs Must Pilot Governance First Before Scaling AI Agents Gartner
- AI data centre regulation just got a template that needs no new law
AI data centre regulation in Pennsylvania now begins with a signature. Before the state will so much as open a developer’s permit file, that developer has to sign a contract accepting a fixed set of conditions and the penalties for breaking them, and persuade the town that has to live with the building to say […] The post AI data centre regulation just got a template that needs no new law appeared first on AI News .
Score: 52🌐 MovesAug 20, 2026https://www.artificialintelligence-news.com/news/ai-data-centre-regulation-pennsylvania-template/ - ChatGPT ability to control iMessage raises Apple privacy concerns
ChatGPT ability to control iMessage raises Apple privacy concerns The Straits Times
Score: 52🌐 MovesAug 20, 2026https://www.straitstimes.com/world/chatgpt-can-now-control-imessage-potentially-raising-apple-privacy-concerns - Anthropic plans to change enterprise data retention policy, source says
Anthropic plans to change enterprise data retention policy, source says Reuters
Score: 52🌐 MovesAug 20, 2026https://www.reuters.com/business/anthropic-plans-change-enterprise-data-retention-policy-source-says-2026-08-20/ - Ex-Google engineer's conviction for stealing AI secrets partially overturned
Ex-Google engineer's conviction for stealing AI secrets partially overturned Reuters
- ‘Buyers aren't yet opening their wallets’: AI-generated assets are flooding online marketplaces, but consumers are snubbing AI for human-made products
‘Buyers aren't yet opening their wallets’: AI-generated assets are flooding online marketplaces, but consumers are snubbing AI for human-made products Fortune
Score: 52🌐 MovesAug 20, 2026https://fortune.com/2026/08/20/ai-product-fatigue-online-marketplace-ecommerce/ - Lifting the floor for all: How Google’s Gemini Models are powering Wadhwani AI’s HealthVaani App for ASHA, Anganwadi workers
The first healthcare touchpoint for more than half of India’s population are the frontline workers such as ASHA and Anganwadi workers. Strengthening their confidence with AI-enabled tools on their hand-held devices, will significantly improve the quality and reach of primary healthcare in the country. Wadhwani AI’s HealthVaani App, powered by Google, is taking the lead in instilling this confidence where it matters the most. Senior Journalist Sonia Singh interacted with Shekar Sivasubramanian - Head, Wadhwani AI and Preeti Lobana - Vice President and Country Manager, Google India. Read on!
- GOP urges top AI firms to do something about the toxic image of data centers
As artificial intelligence data centers continue to proliferate across Ohio and public opposition grows, Axios reported today that the GOP’s Senate campaign arm has privately warned AI companies that they need to do something about the increasingly toxic image of the facilities. The centers have become the battleground for the upcoming election in which Republican […] The post GOP urges top AI firms to do something about the toxic image of data centers appeared first on SiliconANGLE .
- Google Gemini just made it easy to remove watermarks from AI images — here’s how to do it
Google Gemini just made it easy to remove watermarks from AI images — here’s how to do it Tom's Guide
- Welcome to the AI crisis in math
The Verge’s Robert Hart on why new AI discoveries have left the math world 'shell-shocked.’
Score: 52🌐 MovesAug 20, 2026https://www.theverge.com/podcast/982434/ai-math-openai-astra-existential-crisis - Humanoid robots' 'ChatGPT moment' could be 10 years away, Unitree founder says
Unitree founder Wang Xingxing says humanoid robots could take up to 10 years to reach a breakthrough comparable to ChatGPT.
Score: 51🌐 MovesAug 20, 2026https://www.cnbc.com/2026/08/20/unitree-humanoid-robots-chatgpt-moment.html - Cloudera launches Anywhere Cloud to bring cloud-native AI to distributed enterprise data
Cloudera has announced Cloudera Anywhere Cloud, a new hybrid data and AI platform designed to help enterprises build and operate production AI applications across multi-cloud and on-premises environments while retaining control over where their data resides. The post Cloudera launches Anywhere Cloud to bring cloud-native AI to distributed enterprise data appeared first on Express Computer .