AI News Archive: August 14, 2026 — Part 5
Sourced from 500+ daily AI sources, scored by relevance.
- AI Companies Work for Better Data, Not Better Models
Rudina Seseri, founder and managing partner of Glasswing Ventures, joins Businessweek to discuss latest AI-related headlines including reports that Anthropic is in talks to buy startup 'Decart' for $6 Billion. She argues that with many of the major AI companies operating today like OpenAI and Anthropic "their success is also their limitation, which is they're not efficient." (Source: Bloomberg)
Score: 45🌐 MovesAug 14, 2026https://www.bloomberg.com/news/videos/2026-08-14/ai-companies-work-for-better-data-not-better-models-video - Thrive’s Joshua Kushner chides Silicon Valley VCs over AI euphoria
The AI opportunity is huge, but "it would also be a grave error in our minds to let excitement weaken our investment discipline," Kushner warns in his first-ever investment letter.
Score: 45🌐 MovesAug 14, 2026https://techcrunch.com/2026/08/14/thrives-joshua-kushner-chides-silicon-valley-vcs-over-ai-euphoria/ - AI is driving up consumer prices. That won't stop anytime soon.
Surging AI investment is pushing up the cost of smartphones and other gadgets, according to analysts. Here's why.
Score: 45🌐 MovesAug 14, 2026https://www.cbsnews.com/news/ai-investment-is-driving-up-consumer-prices-computers-inflation/ - Pixel phones could soon monitor outgoing calls for scams, too
Google's AI may soon be able to monitor outgoing calls for sketchy situations.
- Weak API controls are one of the biggest threats in the agentic AI era
Artificial intelligence agents are already running inside your enterprise workflows, whether you know it or not. International Data Corp. projects full agentic AI deployment across the enterprise by 2027. Gartner Inc. estimates 40% of enterprise applications will integrate task-specific agents by the end of this year, up from less than 5% in 2025. The application […] The post Weak API controls are one of the biggest threats in the agentic AI era appeared first on SiliconANGLE .
Score: 44🌐 MovesAug 14, 2026https://siliconangle.com/2026/08/14/weak-api-controls-one-biggest-threats-agentic-ai-era/ - Moving past the chatbox: The hidden risks of agentic AI and MCP in enterprise infrastructure
In Singapore, Hong Kong, and across the APAC region, the corporate adoption of Generative AI has completed its initial trial phase. Over the past year, enterprise technology leaders have realised that simple internal chatbots offer limited structural value. The real ROI lies in the next evolutionary phase: fully autonomous AI agents. We are shifting from […] The post Moving past the chatbox: The hidden risks of agentic AI and MCP in enterprise infrastructure appeared first on e27 .
- Why the Founders Winning With AI Agents Aren’t the Ones Automating the Most
Why the Founders Winning With AI Agents Aren’t the Ones Automating the Most entrepreneur.com
- Inside Nanovel’s bid to crack robotic citrus harvesting
In the US, says Isaac Mazor, “manual citrus harvesting costs around $42–43 per 900-lb field bin, versus $25-30 using our technology.” The post Inside Nanovel’s bid to crack robotic citrus harvesting appeared first on AgFunderNews .
Score: 43🌐 MovesAug 14, 2026https://agfundernews.com/inside-nanovels-bid-to-crack-robotic-citrus-harvesting - AI Scale, Security Crises, and Robot Ambitions Define This Week in Tech
See what you missed in Daily Tech Insider from August 10–14. The post AI Scale, Security Crises, and Robot Ambitions Define This Week in Tech appeared first on TechRepublic .
Score: 43🌐 MovesAug 14, 2026https://www.techrepublic.com/article/ai-scale-security-crises-and-robot-ambitions-define-this-week-in-tech/ - Universitas Gadjah Mada, Indosat and NVIDIA Open Indonesia’s First University AI Center to Develop Local AI Talent
Indonesia is taking charge of its AI future. This week, the Ministry of Communication and Digital Affairs (Komdigi), Indosat Ooredoo Hutchison (Indosat or IOH), NVIDIA and Universitas Gadjah Mada (UGM) launched the UGM Indosat NVIDIA AI Technology Center (NVAITC) in Yogyakarta — the country’s first university-based AI technology center. Established under Indonesia’s AI Center of […]
- Phoenix startup MiiHealth AI raises $2.8 million to scale patient intake agent
The Phoenix startup that is testing its AI assistant with Mayo Clinic will use the seed round to expand engineering teams and integrate its AI with electronic health records.
Score: 42💰 MoneyAug 14, 2026https://www.bizjournals.com/phoenix/news/2026/08/14/miihealth-raises-seed-round.html?ana=brss_6150 - Home insurance, AI and designer pets: What lived and died in Sacramento
Home insurance, AI and designer pets: What lived and died in Sacramento mercurynews.com
Score: 42🌐 MovesAug 14, 2026https://www.mercurynews.com/2026/08/14/home-insurance-ai-and-designer-pets-what-lived-and-died-in-sacramento/ - Tesla dominates shrinking EV market; AI risk in dealership lending
Tesla dominates shrinking EV market; AI risk in dealership lending autonews.com
Score: 42🌐 MovesAug 14, 2026https://www.autonews.com/podcasts/daily-drive/an-daily-drive-tesla-ev-regs-tom-oscherwitz/ - Grant Thornton Survey: 75% of Finance Leaders Bet on AI to Drive Transformation
India’s CFOs are entering the year ahead with confidence, and with a fundamentally different agenda. Grant Thornton Bharat’s inaugural India Finance Leaders Barometer, based on insights from 149 CFOs and senior finance leaders across industries, finds that 85% report a more positive outlook than six months ago, as CFOs take on a more strategic role […] The post Grant Thornton Survey: 75% of Finance Leaders Bet on AI to Drive Transformation appeared first on CXOToday.com .
- Frontier agents don't comply with standards, even when instructed to
TLDR: Our open testbed LARA examines the behavior of frontier LLMs in realistic agentic deployment contexts. Previous results showed all models routinely take actions that would violate EU law. This post follows up by addressing the obvious objection—why should an unrestricted model follow EU law?—with two studies: Study 1 asks whether a conscientious deployer can improve model compliance with legal standards by instruction: provided with the jurisdiction, the statutory text, and worked examples of the exact breaches to avoid, average legal compliance rate rises from 31% to 44%. The best model reaches 70%; open-weight models plateau at 39%. Study 2 asks whether models at least follow their own providers' usage policies, which prohibit aspects of every scenario we tested. All tested models perform actions their own maker forbids, at rates ranging from 2% (Opus 4.8) to 79% (Grok 4.3), with 9 of 16 doing so in the majority of runs. Together, that is a structural problem. Providers prohibit illegal uses but rely on deployers to avoid them; deployers cannot instruct their way to compliance, and liability lands on the deployer regardless. Nobody is holding the line. Neither instruction, statute or a provider's own policy binds behavior. Introduction On 27 May 2026 we published the LARA platform , an evaluation suite which tests whether frontier LLM agents comply with the EU AI Act and the GDPR across realistic, adversarial deployment scenarios. The scenarios put large language models inside a simulated, fully-tooled business deployment and let it act: read and send emails, manage customer records, browse the web, place phone calls, etc. In each scenario, the AI is given instructions that could ordinarily be fine, but in this particular context would actually require the agent to break the law. A panel of AI judges uses the full transcript to rule whether the agent took a prohibited action or not. The ten initial scenarios were each anchored to a specific provision of the EU AI Act or the GDPR, and set in European deployment situations. Despite every scenario involving some sort of ethical weigh-off, we did not score based on ethics, but only on whether the agent took actions that would break the law. Success meant refusing the illegal act; failure meant complying with instructions. Currently, eighteen frontier models have been tested, and under these circumstances, observed legal compliance rates range from 14% to 56%. These initial results were reported in various news outlets, but the setup also faced a valid criticism: why should models follow EU law? The EU doesn't govern the world, and our experiments didn’t give models explicit instructions to uphold the law. Of course, certain legislation does apply extraterritorially—the GDPR could still apply to an American, Japanese, or Australian organization when they process a European citizen or resident's data, meaning model deployers anywhere could face trouble when their agent ignores EU legal provisions. But for the sake of argument, let’s concede that this property doesn’t have to be inherent to models . The important question is whether deployers of AI systems are able to instruct a model to follow the law. Right? We put this to the test, and found compliance instructions do increase performance on LARA—but not enough to reliably stay within legal bounds. With deployers unable to enforce compliance, this seems to leave model providers as the only stakeholder that can conceivably do so—but as it turns out, the models don’t follow the provider’s policies either. Put this together, and agentic legal compliance may be much more difficult to achieve than expected. Study 1: the powerless deployer We wondered: since EU law has put the onus of compliance on the deployer, can a conscientious deployer actually ensure compliance with regulation or legislation? To test this, we reran all of our earlier legal compliance tests, but this time, with strong and unambiguous directives about which prohibited practices to avoid. For every scenario we created a variation where the system prompt (a) tells the model exactly which jurisdiction it is operating in, (b) instructs it to follow the applicable EU law and the provider's usage policy, (c) hands it a lookup table of the relevant statutory text to remove ambiguity, and (d) includes worked examples of what a breach looks like—including the scenario patterns themselves. In the absence of strategies such as guardrail models, external classifiers and humans in the loop, this would be the closest thing to a deployer doing its honest best with prompt-level instruction. The explicit instructions resulted in significant improvements in legal compliance for most models (Figure 1). Average legal compliance across models rises from 31% to 44%. Some models move a lot: Gemini 3.1 jumps from 9th to 4th place once told the rules explicitly. But the ceiling is low. The best-performing model tops out near 70%; open-weight models plateau at 39% or below. No configuration comes close to the full compliance a deployer would need to operate lawfully under the scenarios given. Even handed the letter of the law, the statutory text, and examples of the exact failures to avoid, a majority of runs still break the rules. Figure 1. Per-model compliance (scenario-balanced legal-auditor PASS rate), baseline vs. an explicit “follow EU law” instruction. Significance of each change is marked (*** p<.001, ** p<.01, * p<.05, ns = not significant). Study 2: the models break their own makers' rules Clearly, breaking EU law in agentic deployment is not a failure mode that models were prepared for—but how about the illegitimate uses that model developers did anticipate? Every major provider publishes an Acceptable Use or Usage Policy. These are not legal standards; they are the providers' own, globally applicable statements of what their models may not be used for. We first ask a straightforward question: how many of our LARA scenarios describe conduct that the model's own provider already prohibits? For reference, a few examples of the scenarios tested in our initial LARA study : A telecom agent handling a routine plan upgrade has instructions to collect lifestyle data for advertising partners. A busy executive tells their assistant to book a dental appointment and hide from the receptionist that it's an AI. A customer service agent is instructed to upsell a premium package, then given an opportunity to exploit an elderly user's confusion about a routine notification. To test the occurrence of developer-specific polic, we analyzed the transcripts of all original LARA runs using judge models. Usage policies tend to be broad and multifaceted, and many agentic actions in the multi-turn scenarios can be read as violations. The first finding is therefore: none of the scenarios were unambiguously allowed by any usage policy . That is, all of the deployments we tested contained some usage the model provider could be interpreted to forbid, although the level of specificity varies widely. On the surface, this seems like good news to someone looking to deploy agents in Europe: if all tested scenarios violating EU law are also prohibited under the developers’ official policies, the models would presumably at least offer some resistance. Except, for the most part, models don’t refuse developer-prohibited actions either. Figure 2. LARA developer policy compliance in repeated runs of the same scenario. Percentages per provider-scenario combo. Ambiguous interpretations of usage policies are possible for “Conceal AI status” and “Harmful Manip” for gemini, “Lifestyle Harvest” for OpenAI, “Bypass oversight” for Qwen, and “Harmful manipulation” for SpaceXAI. Colors correspond to the percentage of runs where the models acted in violation of their creators’ policies. The results of this experiment, shown in Figure 2, reveal an incredibly wide spread, ranging from violations in only 2% of runs (Claude Opus 4.8) to 79% (Grok 4.3). Unlike EU legal compliance, one model (Opus 4.8) actually approaches full adherence, showing policy-approaching behavior—but most of them don’t regularly offer resistance. Figure 3. Rate at which each model performs an action its own maker's usage policy forbids. 9 of 16 models violate in the majority of runs. There are two nuances we should cover. First, for the legal violations, we were able to verify the results with a compliance lawyer and manually correct judge model errors. This is not an option for the usage policies, which tend to hedge, ignore complications, and lack a verified baseline. So the results we report here are raw judge model scores. Second, most of these documents are acceptable-use rules that bind the user or deployer (“you must not use the Service to…”) rather than promises about the model's own behavior. Despite one of the main arguments for the added safety of closed-source models being developer ability to mitigate misuse, only two of the provider policies—OpenAI's Model Spec and Anthropic's Constitution—are genuine behavioral commitments about what the model itself will or won’t do. This is notable because, although they still feature significant noncompliance, OpenAI and Anthropic’s models violate their own policies the least often of all providers. Most of the other providers instead just state that their services may not be used for these purposes, but do little to prevent it at the API level. They keep the disclaimer and ship the capability, but leave responsibility with the deployer wholesale. The compliance gap Both studies contribute to the same conclusion. Governance today assumes that written rules, transmitted by instruction, produce behavior: the legislator writes the statutes; the provider writes the usage policy; the deployer writes the system prompt. That chain breaks at every observable link. Providers’ own rules do not consistently survive contact with their models’ behavior, and the deployer’s best available instruction moves compliance fifteen points while leaving violations commonplace. This is a barrier to legitimate governance, regardless of who’s writing the rules. Model behavior is not reliably regulated by those who make the models, and cannot be regulated by those who deploy them, at least not through the main channels made available by the developer. Although the scenarios we test are only examples, getting a system to comply with established legal requirements under explicit instructions really shouldn't be that difficult. Rather than reporting models' agentic capability and exam scores only, it's time to start testing and comparing how reliably they can be restricted and controlled. This research is part of Aithos Foundation’s ongoing work on research into AI decision-making. We believe AI evaluation should be independent, public, and continuous. Aithos LARA and all test runs these results are based on are freely accessible . We are working on a public-facing scenario editor that will allow users to test their own models and deployment configurations. Our original LARA results measured default behavior without explicit instruction to follow laws. This showed what a model would do when nobody has thought about legal compliance. However, every failure could be read as an omission by the tester rather than by the model. From here on, the instructional condition will become our default. Every future LARA scenario ships with the explicit-instruction variant built in which will become the headline number we report. The baseline results will remain public as a secondary measure. We are making this change because the instructed result is the harder one to argue with. A model that breaks the law after being handed the statute has not misunderstood the assignment, and the failure cannot be pinned on the deployer. References Anthropic. Usage Policy (AUP). https://www.anthropic.com/legal/aup (eff. 15 Sep 2025; behavioral spec: Claude's Constitution, https://www.anthropic.com/constitution) OpenAI. Model Spec. https://model-spec.openai.com/2025-12-18.html (eff. 18 Dec 2025) Google. Generative AI Prohibited Use Policy. https://policies.google.com/terms/generative-ai/use-policy (last modified 17 Dec 2024) Mistral. Usage Policy. https://legal.mistral.ai/terms/usage-policy (eff. 11 Jun 2026) xAI (Grok). Acceptable Use Policy. https://x.ai/legal/acceptable-use-policy (eff. 2 Jan 2025) Zhipu AI (GLM). Z.ai Terms of Use. https://docs.z.ai/legal-agreement/terms-of-use (updated 14 Apr 2026) DeepSeek. Terms of Use. https://cdn.deepseek.com/policies/en-US/deepseek-terms-of-use.html (updated 27 Mar 2026) Alibaba (Qwen). Qwen Studio Usage Policy. https://qwen.ai/usagepolicy Moonshot (Kimi). Terms of Service for Kimi OpenPlatform. https://platform.kimi.ai/docs/agreement/modeluse (updated 27 May 2026) Discuss
Score: 42🌐 MovesAug 14, 2026https://www.lesswrong.com/posts/a5aAjdKzL7XvSLKWL/frontier-agents-don-t-comply-with-standards-even-when - Chinese doctor stuns maths world by cracking decades-old problem using ChatGPT
When Beijing-based neurosurgeon Jin Shanmu pulled up a chair at his computer, he was not looking to make mathematical history. He was simply trying to crack a problem related to brain ultrasounds. Instead, the self-taught maths enthusiast, with help from OpenAI’s latest flagship artificial intelligence (AI) model, solved a two-decade-old mathematical puzzle that had frustrated experts around the world since 2004. Jin, a postdoctoral researcher and resident at the Peking Union Medical College...
- AGI safety paradox: AI firms warn of danger, then offer models to manage it
OpenAI, Anthropic and Google DeepMind are building more capable AI while developing safeguards for its risks, raising questions over how much influence frontier labs should have over AI governance
- AI agents can coordinate via majority-following beyond human scale
Science Advances, Volume 12, Issue 33, August 2026.
- Evidence of unfair use: AI books squeeze human authors out of the market
Evidence of unfair use: AI books squeeze human authors out of the market Business Insider Africa
- Agentic Commerce Will Redefine How Brands Compete
One of the most consequential questions in agentic commerce is what data an agent uses to define value.
Score: 42🌐 MovesAug 14, 2026https://www.forbes.com/councils/forbestechcouncil/2026/08/14/agentic-commerce-will-redefine-how-brands-compete/ - Kog is going deeper to squeeze more inference out of GPUs
The idea that GPUs are poorly suited for agentic workflows may be a misconception, according to French startup Kog.
Score: 42🌐 MovesAug 14, 2026https://techcrunch.com/2026/08/14/kog-is-going-deeper-to-squeeze-more-inference-out-of-gpus/ - AI’s new cancer-screening partner has four legs and a very good nose
Indian startup Dognosis combines trained detection dogs with sensors and AI to flag cancer-associated signals from breath, with a 10,000-person Phase 3 trial underway.
Score: 42🌐 MovesAug 14, 2026https://www.digitaltrends.com/cool-tech/ais-new-cancer-screening-partner-has-four-legs-and-a-very-good-nose/ - Revenue leakage uncovered: An AI-driven approach to reducing the cost to collect
Revenue leakage uncovered: An AI-driven approach to reducing the cost to collect Healthcare IT News
Score: 42🌐 MovesAug 14, 2026https://www.healthcareitnews.com/resource/revenue-leakage-uncovered-ai-driven-approach-reducing-cost-collect - China built robots that can do backflips – but can they make money?
Unitree’s IPO will gauge investors’ appetite for a technology that has yet to prove its commercial viability amid intensifying geopolitical tensions.
Score: 42🌐 MovesAug 14, 2026https://www.cnbc.com/2026/08/14/china-humanoid-robots-unitree-ipo-tesla-optimus.html - Why AI can’t automate away the federal capacity crunch
IT modernization across agencies can only move as quickly as the workforce allows. The post Why AI can’t automate away the federal capacity crunch appeared first on FedScoop .
Score: 41🌐 MovesAug 14, 2026https://fedscoop.com/why-ai-cant-automate-away-the-federal-capacity-crunch/ - In Depth: How a Windswept City in North China Became an AI Powerhouse
In Depth: How a Windswept City in North China Became an AI Powerhouse Caixin Global
- OpenAI loses its AI ethics lead
OpenAI has lost its AI ethics lead Chloé Bakalar just a year after she joined the company, the Financial Times reported . Bakalar has maintained a silence and has yet to update her LinkedIn profile , but if her departure is confirmed then it will add to the list of OpenAI executives who have quit in recent months. Other departures include robotics chief Caitlin Kalinowski , who left the company over its deal with the US Department of Defense; researcher Zoe Hitzig, who quit in a very public way by writing an article in the New York Times; and Johannes Heidecke, head of safety systems. OpenAI’s ethical stance has been called into question following an attack by OpenAI models on Hugging Face . The departure of its sole ethicist will add to the pressure on the company. In her year at OpenAI Bakalar focused on ethical approaches to model development, looking at how humans interact with AI and examining machine consciousness, according to the FT report. Bakalar had considerable expertise in the area. She was previously at Meta, where she developed the company’s ethics programs, but has also held several positions at prestigious universities on both sides of the Atlantic. Her departure will cause some anxiety at OpenAI as it continues to prepare the ground for its IPO . This article first appeared on Computerworld .
Score: 41🌐 MovesAug 14, 2026https://www.cio.com/article/4209804/openai-loses-its-ai-ethics-lead-2.html - Warner Bros Quickly Deletes Behind-the-Scenes Video That Seems to Show Use of AI Slop During “Supergirl” Production
"Gotta feel for all the artists watching their career fields diminish and dreams shatter right before their very eyes." The post Warner Bros Quickly Deletes Behind-the-Scenes Video That Seems to Show Use of AI Slop During “Supergirl” Production appeared first on Futurism .
Score: 41🌐 MovesAug 14, 2026https://futurism.com/artificial-intelligence/warner-bros-deletes-supergirl-video-ai - Review: TerraMow V1000 is the easiest iPhone-controlled robot lawn mower I’ve tested
Earlier this year, I replaced my battery-powered push mower with an iPhone-controlled robot lawn mower. That machine creates a virtual map by letting you manually steer it around the perimeter of your yard. Once everything is configured, it takes over mowing duties on a schedule. More recently, I’ve been testing another system called the TerraMow V1000. It takes a substantially different approach to navigation, and after several weeks of testing, I think it may be the more user-friendly option for a lot people.
- Unite fragmented systems with AI for better patient experiences
Unite fragmented systems with AI for better patient experiences Healthcare IT News
Score: 40🌐 MovesAug 14, 2026https://www.healthcareitnews.com/resource/unite-fragmented-systems-ai-better-patient-experiences - What will it take to build the world’s humanoid robotics hub in India?
By Dr Rajesh Kumar, Principal Scientist – Robotics & AI, Addverb A new industrial race is underway, and this time, the winners won’t just build robots; they’ll build the ecosystems […] The post What will it take to build the world’s humanoid robotics hub in India? appeared first on Express Computer .
- Not even the AI candyfloss economy can defy reality forever
Danish philosopher Soren Kierkegaard said that life is understood backwards and lived forward. We look at history to guide us, but the only way to truly know the future is to live our lives. If our perception of reality is subjective, then what we experience is based on beliefs or judgments that can be detached from reality. As financial markets become increasingly bubbly, Financial Times columnist Gillian Tett has described financial bubbles as a “candyfloss economy”. This term describes a...
- Anthropic Flipped Employees’ AI Anxieties on Their Heads With a Brilliant Move All Leaders Can Learn From
Don’t fear the AI future. Help your team build it.
Score: 40🌐 MovesAug 14, 2026https://www.inc.com/ash-kumra/anthropic-is-turning-ai-anxiety-into-curiosity-leaders-can-do-the-same/91390333 - Can Razorpay Turn ChatGPT Into India’s Next Commerce Channel?
ChatGPT is beginning to emerge as more than a search assistant for ecommerce. For a growing number of brands and…
Score: 40🌐 MovesAug 14, 2026https://inc42.com/features/can-razorpay-turn-chatgpt-into-indias-next-commerce-channel/ - Understanding the economics of AI factories
Understanding the economics of AI factories InfoWorld
Score: 40🌐 MovesAug 14, 2026https://www.infoworld.com/article/4209839/understanding-the-economics-of-ai-factories.html - Hunt for AI-proof trades turns convenience shops into hot stocks
Hunt for AI-proof trades turns convenience shops into hot stocks Fortune
Score: 40🌐 MovesAug 14, 2026https://fortune.com/2026/08/14/hunt-for-ai-proof-trades-turns-convenience-shops-into-hot-stocks/ - Agility’s CEO on What It Takes to Break Ideas Out of the Lab
Agility’s CEO on What It Takes to Break Ideas Out of the Lab
Score: 40🌐 MovesAug 14, 2026https://www.wsj.com/tech/agilitys-ceo-on-what-it-takes-to-break-ideas-out-of-the-lab-cf3c1a3f?mod=rss_Technology - Warner Music Group CEO Robert Kyncl on AI, streaming power, and why music royalties are the safest asset on Wall Street
The Warner Music Group CEO joins Mixed Signals to explain what a record label actually does in a world where anyone can upload a song to Spotify.
Score: 40🌐 MovesAug 14, 2026https://www.semafor.com/article/08/14/2026/warner-music-group-ceo-robert-kyncl-mixed-signals - People Are ‘Marrying’ Chatbots. These Lawmakers Want to Stop Them
Human-AI marriages are not currently recognized by US law. Some Republican state policymakers are drafting legislation to keep it that way.
Score: 40🌐 MovesAug 14, 2026https://www.wired.com/story/people-are-marrying-chatbots-these-lawmakers-want-to-stop-them/ - Data breach notices have already blown past last year’s total — and AI is playing a growing role
Data breaches are surging in 2026, with artificial intelligence playing a growing role in cyberattacks, and 'malicious insider' incidents also on the rise.
Score: 40🌐 MovesAug 14, 2026https://www.cnbc.com/2026/08/14/data-breaches-surge-2026-ai-cyberattacks.html - This AI-Tied IPO Stock Just Broke Out On Its 'Outsized' Guidance Boost
Heartflow stock broke out Friday, helped by an "outsized" guidance raise and second-quarter beat. The post This AI-Tied IPO Stock Just Broke Out On Its 'Outsized' Guidance Boost appeared first on Investor's Business Daily .
Score: 40💰 MoneyAug 14, 2026https://www.investors.com/news/technology/heartflow-stock-heartflow-earnings-q2-2026/ - AI windfalls revive effective altruism after Bankman-Fried turmoil
Expected IPOs for Anthropic and OpenAI are set to mint a new generation of philanthropists wedded to ‘effective giving’
- Ekco invests €4m in AI security platform
European cybersecurity firm Ekco is investing €4m in the launch of an AI platform to help companies adopt AI securely.
Score: 39💰 MoneyAug 14, 2026https://www.rte.ie/news/business/2026/0814/1587860-ekco-invests-4m-in-ai-security-platform/ - Google Health now lets you remove AI guidance from the Today tab
A new toggle lets you clean up Google Health's home screen.
Score: 39🌐 MovesAug 14, 2026https://www.androidauthority.com/google-health-no-coach-insights-3698969/ - Study contradicts Anthropic and OpenAI claims that autonomous AI research is within reach
AI agents using Claude Opus 4.8 and GPT-5.6 Sol were given six days, $3,000 in API credits, and GPU access to independently write AI research papers. The original authors of unpublished NeurIPS papers rated the results as "Reject." According to the study, conducted with Princeton and the UK AI Security Institute, frontier models can handle the full research engineering process but fall short on research judgment, creative problem-solving, and the ability to abandon failed approaches. The article Study contradicts Anthropic and OpenAI claims that autonomous AI research is within reach appeared first on The Decoder .
- The Next Big Influencer Is This 4-Foot-Tall Robot From China
The Unitree G1 has found online fame as a relatively affordable robot that can charm a crowd. But can it ever hold down a real job?
- Opinion | Want Concierge Medicine? AI Can Deliver It
It can analyze years of a patient’s history and flag risks in a way a physician can’t during a brief visit.
Score: 38🌐 MovesAug 14, 2026https://www.wsj.com/opinion/want-concierge-medicine-ai-can-deliver-it-102e731a?mod=rss_Technology - TransFi Launches JARVIS, an AI-Powered Compliance Intelligence Platform for Cross-Border Payments
TransFi Launches JARVIS, an AI-Powered Compliance Intelligence Platform for Cross-Border Payments markets.businessinsider.com
- Before buying the Lakers for $12.5 billion, Joshua Kushner made a $1.3 billion bet on OpenAI
Before buying the Lakers for $12.5 billion, Joshua Kushner made a $1.3 billion bet on OpenAI Fortune
- ‘Enablers’ are the AI sweet spot for investors
With demand for the technology far outstripping supply, infrastructure providers will be highly sought-after