AI News Archive: July 20, 2026 — Part 5
Sourced from 500+ daily AI sources, scored by relevance.
- 2 years to agentic: Will you comply, or will you grow?
Dubai has put a deadline on autonomous AI. The companies that treat it as a compliance project will end up with a more efficient version of the business they already have. The ones that change the game will end up with a different business. The mandate is real, and it has a clock on it. […] The post 2 years to agentic: Will you comply, or will you grow? appeared first on e27 .
Score: 55🌐 MovesJul 20, 2026https://e27.co/2-years-to-agentic-will-you-comply-or-will-you-grow-20260718/ - Building Governed Agents: A Framework for Cost, Control, and Compliance
Framework for cost-effective, compliant agent governance.
Score: 55🌐 MovesJul 20, 2026https://blog.langchain.dev/blog/building-governed-agents-a-framework-for-cost-control-and-compliance - Robot reboot
A new generation of robotics is powering up. How can leaders capture the value at stake while ensuring successful human–machine collaboration?
Score: 55🌐 MovesJul 20, 2026https://www.mckinsey.com/quarterly/the-five-fifty/five-fifty-robot-reboot - Scaling document classification to 100k+ labels
Across Databricks, thousands of customers build production workloads that map freeform...
Score: 55🌐 MovesJul 20, 2026https://www.databricks.com/blog/scaling-document-classification-100k-labels - Chinese robot makers’ lament: if we only had a better ‘brain’, and more data
Chinese robotics companies lack both sufficient data and a good “brain” to improve the interaction of their products with the physical world, according to industry insiders at the World Artificial Intelligence Conference (WAIC), which concluded on Monday in Shanghai. The most critical challenge for the embodied AI industry was to “link hardware, data, models and real-world scenarios into a closed-loop iterative system”, said Wang Xiaogang, co-founder of SenseTime and chairman of its robotics...
- NVIDIA Wants To Solve India’s GPU Compute Problem, But At What Cost?
Chip giant NVIDIA is all set to change the GPU game across its major markets, and India is no exception.…
Score: 55🌐 MovesJul 20, 2026https://inc42.com/features/nvidia-wants-to-solve-indias-gpu-compute-problem-but-at-what-cost/ - Waymo And Uber Support Bad, Incumbent-Protecting Laws In DC And NJ
Requiring human drivers, 3 sensors, fat permit fees and per-mile taxes are the wrong sorts of regulations, and the companies should avoid endorsing them
- Securing AI Algorithmic Insights
In this report, the authors extend a security framework to provide the first systematic roadmap for protecting algorithmic insights, the novel techniques, methods, and design know-how that materially improve artificial intelligence systems.
- Google’s AI cited Facebook 19.5 million times, new research finds
When Google’s AI answers a question, it is increasingly not reading a website. It is reading a Facebook post. New data suggests social platforms have become a core source for Google’s AI answers. Often the user never visits the platform at all. The findings come from BrightEdge, an enterprise SEO firm. Its tool tracks roughly […] This story continues at The Next Web
Score: 54🌐 MovesJul 20, 2026https://thenextweb.com/news/google-ai-overviews-social-platforms-brightedge - Newman University Announces The Newman Institute for AI and the Common Good
Newman University Announces The Newman Institute for AI and the Common Good USA Today
- As AI Spending Climbs, Enterprises Get Serious About Token Costs
Opaque pricing and backward-looking bills force enterprises to rethink their AI model strategy.
Score: 54🌐 MovesJul 20, 2026https://aibusiness.com/generative-ai/as-ai-spending-climbs-enterprises-serious-about-token-cost - SEBI warns of AI impersonation scams with fraudsters posing as company CXOs
SEBI has warned regulated entities and listed companies about the "Boss Scam", in which fraudsters impersonate senior executives using deepfakes, voice cloning and malware to authorise fund transfers. The post SEBI warns of AI impersonation scams with fraudsters posing as company CXOs appeared first on MEDIANAMA .
Score: 54🌐 MovesJul 20, 2026https://www.medianama.com/2026/07/223-sebi-warns-ai-impersonation-scams-fraudsters-posing-company-cxos/ - Burnham Picks Narayan as First UK AI Minister to Attend Cabinet
New British premier Andy Burnham has named Kanishka Narayan as minister for artificial intelligence, promoting the role to attend the prime minister’s cabinet for the first time.
- Oracle Credit Risk Hits Near 18-Year High on AI Debt Load Angst
The cost of protecting Oracle Corp.’s debt against default reached a fresh multi-year high on Monday while its existing bonds sold off, as doubts grew over whether the company’s massive investments in artificial intelligence will pay off.
- The Canadian companies trying to break Palantir’s grip on intelligence software
The post The Canadian companies trying to break Palantir’s grip on intelligence software appeared first on The Logic .
- Building the network for agentic AI: The foundation for autonomous enterprise operations
Enterprise AI is entering a new phase. While the first wave of generative AI focused on human productivity and content creation, the next wave — agentic AI — will fundamentally change how organizations operate. Agentic AI systems are capable of reasoning, planning, making decisions and executing actions across applications, workflows and business processes with minimal human intervention. As organizations move toward agentic frameworks that can independently resolve customer issues, optimize supply chains, manage infrastructure, coordinate workflows and even operate IT environments, one reality becomes clear: The network becomes the nervous system of the autonomous enterprise. The infrastructure requirements of agentic AI differ dramatically from those of traditional applications. These systems are highly distributed, continuously exchanging information, interacting with APIs, accessing multiple data sources and making decisions in real time. The performance, security, visibility and adaptability of the network will directly determine the effectiveness of AI agents. Organizations that view AI readiness solely as a compute or data challenge risk overlooking one of the most critical enablers of future success — the network itself. From AI-ready networks to autonomous networks The long-term destination is the autonomous network : A network capable of self-monitoring, self-optimizing, self-healing and self-securing through the use of AI and automation. However, autonomous networking will not emerge overnight. The investments enterprises make today to support agentic AI are the same foundational building blocks required for tomorrow’s autonomous operations. In many ways, agentic AI serves as both the driver and beneficiary of network transformation. AI agents require networks that can dynamically adapt to changing demands, while autonomous networks will increasingly rely on AI agents to manage and optimize themselves. The result is a reinforcing cycle where AI and networking evolve together. The core characteristics of the network of the future One of the most critical requirements for AI-ready networks is real-time observability and telemetry. Agentic AI thrives on context, and AI agents must continuously gather information from users, applications, devices, clouds, security systems and operational platforms. Future-ready networks must provide end-to-end visibility across campus, branch, cloud and data center environments. High-fidelity telemetry streams, real-time performance monitoring, application-aware analytics, AI-aware analytics and unified operational visibility are essential. Without comprehensive visibility, AI agents operate with incomplete information, limiting their effectiveness and increasing operational risk. Another cornerstone is intent-based automation. Traditional networks are configured manually, often requiring administrators to define thousands of individual settings. In contrast, autonomous networks operate according to business intent. Enterprises increasingly need to define desired outcomes — such as maintaining application performance, optimizing user experience or automatically isolating compromised devices — rather than micromanaging configurations. The network continuously adjusts itself to achieve those objectives, providing the foundation upon which AI agents can make decisions safely and consistently. Agentic AI also introduces entirely new traffic patterns that require AI-optimized connectivity. Large language models, retrieval systems, vector databases, cloud AI services, edge inference platforms and multi-agent orchestration frameworks create significant east-west and cloud-bound traffic. Future networks must provide low-latency connectivity, high-capacity fabrics, dynamic traffic engineering, edge-to-cloud optimization and policies that identify and prioritize AI workloads. The organizations that can move data efficiently will gain a competitive advantage in AI execution speed and responsiveness. Security is another non-negotiable element. Agentic AI expands the enterprise attack surface because AI agents increasingly access sensitive systems, interact with APIs, consume proprietary data and execute actions across business environments. Future-ready networks must embed zero trust security into their architecture, with continuous identity verification, fine-grained access controls, microsegmentation, policy-driven authorization and continuous risk assessment. Security can no longer be bolted onto the network; it must be integral to its design and AI agents need to adhere to their own identity rules. Finally, distributed intelligence across edge and cloud environments is essential. Many AI use cases require decisions to occur close to the source of data. Manufacturing systems, healthcare environments, retail operations, transportation networks and smart facilities often cannot tolerate the latency associated with centralized processing. Future networks must support edge AI deployment, distributed processing architectures, local inference, hybrid cloud operations and intelligent workload placement. The ability to move intelligence closer to users, devices and operational environments will become increasingly important as agentic AI expands across the enterprise. Human expertise remains essential Despite rapid advances in AI, the future will not eliminate the need for human expertise. In fact, it may increase its importance. One of the most significant misconceptions surrounding AI is that automation eliminates the need for skilled professionals. The reality is that autonomous systems require expert oversight, governance, validation and continuous optimization. As AI systems become more capable, enterprises will need professionals who understand network architecture, security policy, AI governance, operational risk management, data quality, regulatory compliance and human-in-the-loop decision frameworks. The challenge is compounded by the unprecedented pace of AI innovation. New models, architectures, orchestration frameworks, security concerns and governance requirements emerge almost monthly. Most enterprise IT teams cannot be expected to independently evaluate every development while simultaneously modernizing infrastructure and maintaining day-to-day operations. Organizations need access to experts who continuously track technology evolution, understand emerging best practices and can help translate innovation into practical deployment strategies. These experts provide not only implementation support but also ongoing operational guidance, helping enterprises maintain appropriate human oversight as AI capabilities expand. The future is not fully autonomous decision-making without people; it is intelligent automation operating under expert human governance. 5 actions enterprises should take now Organizations should be preparing for the autonomous future right now. The following investments deliver immediate value while laying the groundwork for long-term AI transformation: Modernize network observability. Establish comprehensive visibility across users, applications, devices, clouds and infrastructure. Rich telemetry and operational data will become the fuel that powers both Agentic AI and autonomous network operations. Build an automation-first operating model. Identify repetitive operational processes and begin automating them. Automation maturity is a prerequisite for autonomous networking and creates the operational foundation AI agents will eventually leverage. Adopt zero-trust principles across the enterprise. Implement identity-centric security controls, segmentation and continuous policy enforcement. As AI agents gain access to enterprise systems, security architectures must evolve to leverage the same identity controls. Design for edge-to-cloud AI workloads. Evaluate network architectures for latency, bandwidth and resiliency requirements associated with distributed AI. Future AI deployments will span data centers, public clouds, branch locations and edge environments. Invest in skills and strategic partnerships. Develop internal expertise while leveraging partners that possess deep networking, automation, security and AI knowledge. Human expertise remains one of the most important success factors in building AI-ready and autonomous infrastructures. The road ahead Agentic AI is poised to transform enterprise operations in much the same way cloud computing transformed infrastructure and the internet transformed business itself. But AI agents cannot operate effectively without a modern network foundation. The enterprises that succeed will recognize that AI readiness extends beyond models and data. It requires networks that are observable, automated, secure, intelligent and increasingly autonomous. The investments made today in AI-ready networking are not merely infrastructure upgrades — they are strategic building blocks toward the autonomous enterprise of the future, where AI agents and autonomous networks work together under human guidance to deliver unprecedented levels of agility, efficiency, and innovation. This article is published as part of the Foundry Expert Contributor Network. Want to join?
- How Florida State University Is Building Its AI Ecosystem
Faculty and students, academic researchers, the staff and administrators working behind the scenes at colleges and universities around the country are all leveraging artificial intelligence to teach, learn, and work more efficiently and effectively. One recent survey, in fact, found that 95% of students and educators are actively using AI tools. And while the surge in AI adoption in higher education kicked off in 2022 with OpenAI’s release of ChatGPT, today it’s being driven by a wide range of platforms from established and emerging companies alike. In this environment, a growing number of…
Score: 53🌐 MovesJul 20, 2026https://edtechmagazine.com/higher/article/2026/07/how-florida-state-university-building-its-ai-ecosystem - Armatrix builds snake-like robots so humans don't have to crawl into dangerous tanks
Armatrix builds snake-like robots so humans don't have to crawl into dangerous tanks YourStory.com
Score: 53🌐 MovesJul 20, 2026https://yourstory.com/2026/07/armatrix-snake-robots-dangerous-tank-inspections - Samsung’s secret AI chip could finally cool down Exynos phones
A laptop-bound AI chip from Samsung's Exynos design team hints at a real fix for the thermal and battery problems.
Score: 53🌐 MovesJul 20, 2026https://www.digitaltrends.com/computing/samsungs-secret-ai-chip-could-finally-cool-down-exynos-phones/ - Inference startup Infinity raises $15M from Touring Capital, OpenAI and Anthropic researchers
AI infrastructure company Infinity announced Monday a $15 million raise at a $100 million valuation from investors including Touring Capital, Principal VC, and researchers from companies such as OpenAI and Anthropic.
- 'For humans, the capacity to say, 'I don't know,' is very important': Report finds AI really might be harming our critical thinking skills
New paper reveals human critical thinking could be declining as participants choose to trust AI-generated outputs too much.
- Chinese Companies Need a New Cloud Playbook for AI Abroad
Chinese Companies Need a New Cloud Playbook for AI Abroad Caixin Global
Score: 52🌐 MovesJul 20, 2026https://www.caixinglobal.com/2026-07-20/chinese-companies-need-a-new-cloud-playbook-for-ai-abroad-102466187.html - Enlightenment Capital backs D.C.-based Quorum, a AI firm for government affairs
The Chevy Chase investment firm led by Devin Talbott now has $2.1 billion in assets under management.
Score: 52💰 MoneyJul 20, 2026https://www.bizjournals.com/washington/news/2026/07/20/enlightenment-capital-quorum-analytics.html?ana=brss_6150 - Former Microsoft AI Leaders Are Spending $1M To Prove AI Can Replace CEOs
Skyfall AI, from the team behind Microsoft's $160 million Maluuba acquisition, will buy a SaaS company and let AI run it as CEO.
- Te Whatu Ora taps NHS partner for AI development
Te Whatu Ora taps NHS partner for AI development Healthcare IT News
Score: 52🌐 MovesJul 20, 2026https://www.healthcareitnews.com/news/anz/te-whatu-ora-taps-nhs-partner-ai-development - On theCUBE Pod: IBM’s AI test, Nvidia’s lead and the race for enterprise intelligence
Artificial intelligence companies are racing to control the infrastructure, data and software layers that will power enterprise intelligence. Nvidia Corp. remains far ahead in accelerated computing, but Advanced Micro Devices Inc., Broadcom Inc. and other challengers are positioning themselves for a market in which demand may support multiple winners. At the same time, IBM Corp. […] The post On theCUBE Pod: IBM’s AI test, Nvidia’s lead and the race for enterprise intelligence appeared first on SiliconANGLE .
Score: 52🌐 MovesJul 20, 2026https://siliconangle.com/2026/07/20/enterprise-intelligence-ai-race-thecubepod/ - 'Ouroboros' AI system changes what edge devices see video in real time
A research team led by Kyunghan Lee, a professor in the Department of Electrical and Computer Engineering at Seoul National University College of Engineering, has developed an artificial intelligence system, "Ouroboros," that performs vision transformer-based video analysis 2.61 times faster using 13.0% of the computation required by conventional methods on edge devices.
Score: 52🌐 MovesJul 20, 2026https://techxplore.com/news/2026-07-ouroboros-ai-edge-devices-video.html - How is AI changing weather forecasting?
How is AI changing weather forecasting? wadham.ox.ac.uk
- MEDIA ALERT: VanishID to Present Agentic AI Platform for State Governments at NCSL Legislative Summit
MEDIA ALERT: VanishID to Present Agentic AI Platform for State Governments at NCSL Legislative Summit azcentral.com and The Arizona Republic
- The Massive Supply Deals Feeding the AI Frenzy Are No Sure Thing
Long-term arrangements touted by companies like SK Hynix aren’t as solid as they seem.
- The Self-Driving Company Is AI’s Next Business Model.
To understand future AI-native firms, businesses must study companies already operating themselves, just as military strategists learn from Ukraine's drone warfare.
Score: 51🌐 MovesJul 20, 2026https://www.forbes.com/sites/johnsviokla/2026/07/20/the-self-driving-company-is-ais-next-business-model/ - Apple could ‘run the table’ on AI if it does things right
Looking ahead just a short time, Apple could hold a powerful position in AI where it most makes sense: deployment. Not only will the company offer up its own AI models for the kind of tasks millions use ChatGPT to do today, but it will provide more sophisticated on-device agentic models to help users get things done through Siri AI. Apple also offers limited capacity for more complex tasks through Private Cloud Compute , and, in partnership with the likes of Google in the US and Alibaba in China, the company is giving users a trusted conduit through which to access even more sophisticated AI services. Deeply deployable Critics can say it took Apple a long time to get to this point, but they also seem to think the company has finally got the mix right with its series 27 operating systems. Arriving late to a party doesn’t mean you won’t shine once you get there . Apple is also coming up the inside lane around frontier AI, with iterative OS and hardware enhancements that mean its devices become increasingly effective for Edge AI use cases , on device — no cloud service required. The company appears to be digging down into those use cases. Mark Gurman at Bloomberg recently predicted that future M7 Ultra Macs will support as much as 1.5TB RAM, making these systems more than capable of running full weight frontier models in people’s offices, colleges, and homes. While that does assume the AI-flationary memory market can supply that much RAM at prices humans can afford, it is also true that people are already running AI clusters using off-the-shelf Mac minis networked over Thunderbolt cables. It’s no stretch to believe this will continue to be the case , and that it will even broaden as the power/performance offered at the high end grows. What’s wrong with good enough? When combined with open AI stacks, particularly newly emerging varieties, Apple’s platforms should become leading contenders for private AI services and edge AI. Many business users will leap at the chance to offer their workers powerful, self-hosted, private AI services using one or more daisy-chained Mac Studios or Mac minis. The recent craze in deployment of both Macs to support OpenClaw instances shows they already are. Ultimately, these different slices of momentum mean I agree with investor Jason Calacanis that Apple is in position to apply a great deal of pressure on OpenAI and Claude just by putting models on their devices. It’s also worth thinking about how people use AI today. How many of the queries made in the world right now constitute relatively simple tasks that could be transacted by on-device AI, such as the emerging new version of Apple Intelligence or even smaller LLM models running on device? You can even run PrismML’s 1-bit, 27-billion parameter Bonsai on an iPad using the Locally app, and that’s in the here and now. What happens? Pretty soon you’ll find people recognize that they can already run the vast majority of their AI-augmented workflows using services they have on their existing device or can access on their on-prem Mac set-ups. And, of course, as people get used to running small tasks locally and larger tasks on premises, the actual space in which they need to turn to cloud-based frontier models will erode . That’s even as companies like PrismML work towards slimming down full-weight models so they don’t need to run on a server at all. “It’s going to be wild when people have unlimited tokens on their desks,” said Calacanis in a podcast round table discussion. Who has the most to lose? The current incarnations of AI felt like they came from nowhere. Most people weren’t aware of the technology until returning to work after the 2022 holiday season. Since then, the industry has proliferated with dozens of competing models, most recently including powerful but affordable frontier models such as Qwen and Kimi.ai. These models aren’t necessarily all as good as one another, but in many cases for much of what we do, we’ll find them to be good enough. That’s an existential crisis for some, as industry observers now think the inevitable pricing pressure means some services might have over-invested in capacity before finding any way to turn a profit. Those profit-seeking services are the ones with the most to lose as Apple extends its hardware advantage, democratizing AI access for all while providing platforms suitable for edge AI, on-premises AI, private AI, and even AI access using third-party services. (The need for the latter will shrink as the capabilities of the former get better.) Cupertino rising What does this all mean? While the industry remains young, it is already fragmenting. And striding through the dust of that process comes Apple, equipped with the hardware, software, and approach to build its business even as the enterprise of first mover AI services erodes. You can follow me on social media! Join me on BlueSky , LinkedIn , Mastodon and subscribe to my daily Apple-related news summaries at The Core .
Score: 51🌐 MovesJul 20, 2026https://www.computerworld.com/article/4198808/apple-could-run-the-table-on-ai-if-it-does-things-right.html - Dr. Jill Lepore on why AI backlash is vital for the future
The Artificial State author explains how counting everything breaks society — and how college grads booing AI might save us all.
Score: 50🌐 MovesJul 20, 2026https://www.theverge.com/podcast/967884/jill-lepore-history-ai-artificial-state-elon-musk - XMPro Named as a Sample Vendor for Agentic AI category in the Gartner® Hype Cycle™ for AI in Manufacturing, 2026
XMPro Named as a Sample Vendor for Agentic AI category in the Gartner® Hype Cycle™ for AI in Manufacturing, 2026 azcentral.com and The Arizona Republic
- How our universal content processing platform Riviera evolved for AI and beyond
Riviera is the Dropbox content processing platform that’s been iteratively improving content transformation in our products for roughly a decade.
- Don't Want to Give AI Agents Your Logins? Here's How to Avoid It With 1Password for Claude
Don't Want to Give AI Agents Your Logins? Here's How to Avoid It With 1Password for Claude PCMag
Score: 50🌐 MovesJul 20, 2026https://www.pcmag.com/news/giving-an-ai-agent-your-password-heres-how-1password-means-you-dont-have - NVIDIA data center hardware is being cooled with water 'hotter than a hot tub'
It may seem counterintuitive, but 113-degree water is still cool enough to cool NVIDA's latest hardware.
Score: 50🌐 MovesJul 20, 2026https://www.engadget.com/2214770/nvidia-data-center-liquid-cooling-temperature-heat/ - How the ‘fast and furious’ growth of AI data centres has become a Toronto-area election issue
How the ‘fast and furious’ growth of AI data centres has become a Toronto-area election issue Toronto Star
- Should AI investing involve diversifying into Chinese tech stocks to counterbalance new industry rivals?
Artificial intelligence innovators outside of the United States can no longer be ignored. Does it mean that your AI-focused portfolio is at risk if you haven’t looked beyond Wall Street? There’s no doubt that AI is the largest driver of growth for investors in the post-pandemic landscape, but as valuations continue to swell, challenges to […] The post Should AI investing involve diversifying into Chinese tech stocks to counterbalance new industry rivals? appeared first on e27 .
- The cost of intelligence: How CIOs can manage AI demand at scale
As AI costs spiral, CIOs need to manage enterprise AI demand to optimize for outcomes, not just cost.
- Large language models in the UK: public use, trust, and attitudes - ORA
Large language models in the UK: public use, trust, and attitudes ORA - Oxford University Research Archive
- No longer token economy? SenseTime bets on ‘task economy’ as token prices set to drop
The commercialisation of artificial intelligence is poised to transition from a “token economy” to a “task economy” as token prices decline in the coming years, according to Chinese AI company SenseTime. “The pricing of tokens will inevitably drop, much like the cost of telecoms data did two decades ago,” SenseTime CEO Xu Li said in an interview with the South China Morning Post. After foundational models and computing power become basic infrastructure, the true commercial value will shift...
- Salesforce VP on the leaky AI pipeline: why cheaper tokens won’t fix enterprise AI
Salesforce VP on the leaky AI pipeline: why cheaper tokens won’t fix enterprise AI Fortune
Score: 50🌐 MovesJul 20, 2026https://fortune.com/2026/07/20/leaky-ai-pipeline-cheaper-tokens-salesforce-vp/ - Qwen 3.8 ✨, Kimi Code CLI 👨💻, Netflix’s LLM stack 🤖
Qwen 3.8 ✨, Kimi Code CLI 👨💻, Netflix’s LLM stack 🤖
- How the Texas data center boom is pushing Austin to new limits on development
How the Texas data center boom is pushing Austin to new limits on development Austin American-Statesman
- Luciana Womenswear Introduces Nigeria’s First AI Virtual Try-On, Setting a New Standard for Online Fashion Retail
Premium womenswear brand Luciana launches Nigeria's first artificial intelligence-powered virtual...
Score: 50🌐 MovesJul 20, 2026https://techpoint.africa/brandpress/luciana-womenswear-introduces-nigerias-first-ai-virtual-try-on/ - Spain’s female-led FoodTech startup Sensesbit raises €1 million to scale its AI-powered sensory intelligence platform
Sensesbit, a Lugo-based FoodTech startup specialising in advanced sensory analysis, has closed a €1 million funding round to accelerate its international expansion. The round included participation from Clave Capital, Eoniq Fund, WindOne Consultores, and Paraíso Natural Ventures, the investment vehicle promoted by Grupo Central Lechera Asturiana. Maruxa Quiroga, CEO and co-founder of Sensesbit, said, “For […] The post Spain’s female-led FoodTech startup Sensesbit raises €1 million to scale its AI-powered sensory intelligence platform appeared first on EU-Startups .
- 7 issues impacting AI strategies — and how CIOs should respond
CIOs remain at the forefront of setting the course for AI adoption in their organizations. In fact, 82% of CIO respondents to CIO.com’s 2026 State of the CIO survey are responsible for researching and evaluating AI products, with 78% of IT leaders saying their IT departments are driving AI adoption efforts, with business units aligning their strategies accordingly. As such, CIOs are leading or co-leading AI strategies at the majority of organizations, with many also playing a key role in tackling AI change management . They report encountering numerous factors — from heightened pressure to deliver ROI to challenges with trust in AI outputs — as they formulate and shape those AI strategies. Here’s a look at seven notable issues impacting AI strategies in 2026. 1. Increasing pressure to show ROI for AI investments The era of AI experimentation and pilots is over. Boards and CEOs are making it clear they want to see quantifiable returns from their AI investments . Kyndryl’s 2025 Readiness Report , for example, found that 61% of senior business leaders and decision-makers felt more pressure to prove ROI on their AI investments than they had the prior year. “The era of funding AI is shifting from everything all-in to every project has to have line of sight to some financial value at the end of the day. It’s moving from the experimentation phase to expecting measurable outcomes,” says Jim Piazza , chief AI officer at IT services firm Ensono. As a result, Piazza says companies, both his own as well as those he advises, are more diligent about building business cases that estimate implementation costs, AI run costs, and expected benefits so they’re primed to pursue AI initiatives that will deliver ROI. That strategy seems to be paying off. According to the May 2026 AI Momentum Survey from Dun & Bradstreet , 67% of 10,000 businesses surveyed reported seeing early signs or pockets of ROI, 20% reported multiple projects delivering ROI, and 10% reported strong ROI. That’s a big jump from earlier surveys that found few AI initiatives providing returns. For example, PwC’s 2026 Global CEO Survey , released in January, found that 56% of CEOs saw no significant financial benefit from AI to date, while The GenAI Divide: State of AI in Business 2025 from MIT found that 95% of enterprise generative AI projects failed to show measurable financial returns within six months. 2. The need to harness AI for transformation The No. 1 concern for CEOs this year, according to PwC’s 2026 Global CEO Survey , is whether they’re transforming fast enough to keep pace with technological change, cited by 42% of respondents as their top concern. And 68% of the 1,120-plus C-suite executives surveyed by KPMG for its May 2026 Adaptability Pulse Survey said they feel pressure to accelerate innovation. That in turn is influencing AI strategies. Steve Santana , CIO and head of AI at ETS, the world’s largest private nonprofit educational testing and assessment organization, says his company is “pivoting from working on enterprise efficiencies using AI to figuring out how to deliver assessments,” adding that “AI will enable innovation we couldn’t get to before.” For ETS, that means reimagining how the company delivers its core products, “finding areas to do something you couldn’t do before because it was too big or too daunting,” such as having more interactive tests and assessments at scale, Santana says. And while Santana believes organizations can’t move too slowly, he predicts innovation will trump speed. “The winners and losers in the AI race aren’t always going to be the ones that got there the fastest,” he says, observing that those who move too fast “can drive behaviors that are very dangerous.” He adds, “I’m not advocating for moving slow; I’m advocating moving at pace. It’s better to be measured in your approach.” 3. The black box of AI costs CIOs are struggling to calculate the full cost to run AI for their use cases, with estimates coming in well under what their actual bills will be. Consider the figures from research firm IDC, which found that global 1,000 companies will underestimate their AI infrastructure costs by 30% through 2027 . That makes identifying which AI use cases will produce quantifiable value much more challenging, which in turn makes determining a winning AI strategy harder to do. CIOs, however, say they can’t let that stop them from advising their C-suite colleagues on which AI use cases are likely to be winners. “You can’t sit on the sidelines and wait and watch. The general conclusion is you’re going to lose if you do that, so you have to play even though the cost dynamics are not really well understood,” says Mohan Sankararaman , executive vice president and CIO of First Horizon Bank. Sankararaman says he’s devising his AI strategy with that uncertainty in mind. “It’s up to me and my team to figure out how to optimize our use for costs, just like we did with cloud,” he says, noting that part of his strategy is to avoid infrastructure choices that could result in AI vendor lock-in and, thus, getting stuck with that vendor’s bills. “IT has to get the engineering right and not overengineer solutions to make sure the AI strategy we pursue delivers returns,” he adds. Researchers recommend such approaches. In a blog highlighting the IDC research , Jevin Jensen, research vice president for infrastructure and operations at IDC, wrote that “organizations successfully navigating this challenge are ones that effectively share a common trait: they’ve reimagined FinOps as a strategic team, not an after-the-fact accounting exercise. They treat AI economics as a living ecosystem — measurable, visible, and continuously optimized.” 4. Aligning use cases to business strategy There are an overwhelming number of potential use cases, so execs must pick and prioritize those that will help them achieve their strategic goals. That’s easier said than done. Enterprise Strategy Group’s 2025 report on generative AI’s ROI surveyed 1,900 business and IT leaders across nine countries and found that 71% had more potential use cases that they want to pursue than they can possibly fund; 54% said selecting the right use cases based on objective measures like cost, business impact, and the organization’s ability to execute is hard; and 71% acknowledged that selecting the wrong use cases will hurt their company’s market position. Furthermore, 59% of respondents said advocating for the wrong use cases could cost them their job. Longtime CIO adviser Larry Wolff says challenges picking and prioritizing use cases stems in part from boards and CEOs commanding their teams “to do AI.” Such directives, he explains, puts the technology first and business goals second — something CIOs have been trying to avoid for years. “There should not be a technology strategy. There should be a business strategy with a technology component. The same applies to AI,” says Wolff, now CIO of Preferred Travel Group. “We need to talk about business challenges and opportunities first and then talk about how AI can solve for those.” 5. Human readiness to use AI Even as Sankararaman and his executive colleagues build the bank’s AI strategy, he still sees the need to improve the organization’s understanding of the technology . “Everybody has a basic understanding, but AI fluency isn’t where it should be,” he says, noting that a subpar level of fluency “can hamper creativity.” “If the strategy is to become top notch in, say, customer experience, we have to determine how to achieve that. And if you start building the road map but you don’t know what the technology can do, then the strategy will be limited,” he adds. Sankararaman considers running AI boot camps for executives and their direct reports to improve their knowledge of AI and its transformative capabilities. “Not everyone needs to be an AI expert, but we still need to have a level of understanding of, say, what a large language model is and how to apply it and other elementary things like that. The hope is that when we do talk about strategy for business outcomes, everyone will know how to leverage AI,” he explains. According to Jamaal Justice , principal for people consulting at EY, concern about AI fluency is widespread. “One of the biggest challenges that impacts the success of an AI strategy is human readiness,” Justice says. He points to EY research showing “that while 88% of employees use AI at work, only 28% of organizations have positioned employees to achieve transformative business impact from AI. This underscores that the challenge is not access, but adoption and readiness.” Like Sankararaman, Justice acknowledges that it’s OK to have a spectrum of knowledge and use among workers. But success with AI “depends on aligning mindsets, skillsets, and toolsets, by creating the right conditions for both workforce readiness and effective technology use,” he says. “Organizations that integrate human capability with technology and fundamentally rearchitect work using a human-centered and value-oriented approach will unlock value at scale,” he adds. “Those that don’t risk fragmented adoption and limited returns.” EY research confirms as much, finding that productivity gains can fall by more than 40% when AI is deployed on weak talent foundations, including poor learning, culture, and incentives. 6. Data readiness for AI use Data readiness is also lagging at most organizations, further hindering AI ambitions. According to a 2026 report from Cloudera and Harvard Business Review Analytic Services titled Taming the Complexity of AI Data Readiness , 73% of surveyed business leaders said their organization struggles with AI data preparation. The top obstacles are siloed data and difficulty integrating data sources (56%), lack of a clear data strategy (44%), data quality and bias issues (41%), and regulatory constraints on data use (34%). To ensure AI success, “a radical reshaping of the data landscape is needed,” says Steve Prewitt , who as chief data and AI officer at IT services firm Genpact advises clients on AI deployments for their own organizations. That reshaping is more critical today as agentic AI becomes more prevalent, Prewitt observes. Organizations need high-quality well-governed data to enable and trust AI agents to make real-time decisions autonomously. Otherwise, organizations either can’t move forward with deploying agents or, if they do, risk triggering cascading failures. 7. Engendering trust ETS CIO Santana and his colleagues recognize AI’s potential to deliver faulty outputs, whether from problematic data, drift, or other problems. Everyday users recognize that potential, too. That’s why the issue of trust has a significant impact on the nonprofit’s AI strategy. Companies such as ETS that provide critical, high-stakes services know they must earn trust by building AI use cases that can consistently and demonstratively deliver accurate outputs, Santana says. ETS’s strategy is to highlight where AI is making high-stakes decisions and to detail what steps the company must take to ensure that it consistently delivers accurate, trustworthy outputs and that it conforms to established standards and requirements, he says. “You don’t want someone to feel the results may be wrong if you’re using AI to assess a person and their future depends on it,” he notes. “You want to remove any doubts [in such AI use cases], and the strategy should ensure that. The strategy should include all the work needed to have that trust.”
Score: 49🌐 MovesJul 20, 2026https://www.cio.com/article/4198030/7-issues-impacting-ai-strategies-and-how-cios-should-respond.html - Legal AI’s Legibility Problem
Increasing reliance on language models for legal tasks have led to calls for more information about the likelihood and severity of potential errors. Realizing useful and robust transparency requires a
- The human capabilities separating AI leaders from AI followers
What does effective, measurable human-AI collaboration really look like?
Score: 49🌐 MovesJul 20, 2026https://www.hrdive.com/spons/the-human-capabilities-separating-ai-leaders-from-ai-followers/825324/