AI News Archive: June 12, 2026 — Part 5
Sourced from 500+ daily AI sources, scored by relevance.
- When reality is in doubt, news editors ask this Berkeley professor: Is it AI?
When reality is in doubt, news editors ask this Berkeley professor: Is it AI? San Francisco Chronicle
Score: 52🌐 MovesJun 12, 2026https://www.sfchronicle.com/bayarea/article/hany-farid-ai-deepfakes-22291920.php - Understanding AI with Mark Daley: Canada's new AI strategy
Understanding AI with Mark Daley: Canada's new AI strategy CBC
- AI Board convenes its eighth meeting
AI Board discusses European AI strategy and AI Act implementation.
Score: 52🌐 MovesJun 12, 2026https://digital-strategy.ec.europa.eu/en/news/ai-board-convenes-its-eighth-meeting - AI agents can decide when to strike deals using UPI mandate
Pine Labs introduces agent-to-agent payments in India. This allows AI agents to make payments autonomously. Users set mandates, and AI agents handle transactions based on conditions. This innovation removes the need for human checks at payment points. Partnerships with Gullak and Vijay Sales demonstrate its potential for automated purchases.
- New OpenAI Academy courses for the next era of work
OpenAI introduces three Academy courses that help people build practical AI skills, create repeatable workflows, and apply agents in everyday work.
- The AI industry's platform trap is starting to look a lot like Microsoft's
Anthropic is throttling its new Mythos model for certain tasks while building apps that directly compete with its largest customers. Customers, partners, and investors are pushing back. The article The AI industry's platform trap is starting to look a lot like Microsoft's appeared first on The Decoder .
Score: 52🌐 MovesJun 12, 2026https://the-decoder.com/the-ai-industrys-platform-trap-is-starting-to-look-a-lot-like-microsofts/ - TensorWave COO Piotr Tomasik sees 'dynamic change' in AI investment
TensorWave COO Piotr Tomasik sees 'dynamic change' in AI investment
Score: 52🌐 MovesJun 12, 2026https://qz.com/tensorwave-coo-piotr-tomasik-ai-investment-dynamic-change - TCS and Opendoor fuel worries over AI impacting India's tech jobs
TCS and Opendoor fuel worries over AI impacting India's tech jobs Nikkei Asia
- 40% of enterprises will scrap AI agents - 3 ways to ensure yours don't fail
How do you create real ROI from autonomous AI? Three digital leaders share lessons they've learned in the field.
Score: 51🌐 MovesJun 12, 2026https://www.zdnet.com/article/enterprises-scrapping-ai-agents-how-to-ensure-yours-dont-fail/ - Video Friday: Robotic Motion Discovery Reveals Unusual Behaviors
Your weekly selection of awesome robot videos
- Waymo Premier — Ah, This Is Where The Company’s Headed!
Waymo has long been considered the US leader in self-driving vehicles, and parent Google before it when Waymo wasn’t yet a brand. Well, I guess there are many people who have considered Tesla the leader for the past several years, but that is based on the assumption that Tesla’s approach ... [continued] The post Waymo Premier — Ah, This Is Where The Company’s Headed! appeared first on CleanTechnica .
Score: 50🌐 MovesJun 12, 2026https://cleantechnica.com/2026/06/11/waymo-premier-ah-this-is-where-the-companys-headed/ - Can Ai Have Intuition About Critical Decisions? AngelAi's new AI model with Uncertainty Awareness avoids bad decisions
Can Ai Have Intuition About Critical Decisions? AngelAi's new AI model with Uncertainty Awareness avoids bad decisions azcentral.com and The Arizona Republic
- In aerospace, AI isn’t replacing workers. It’s filling a shortage
In aerospace, AI isn’t replacing workers. It’s filling a shortage SpaceNews
Score: 50🌐 MovesJun 12, 2026https://spacenews.com/in-aerospace-ai-isnt-replacing-workers-its-filling-a-shortage/ - Q&A: Can we trust AI models? Researchers explore the roots of chatbot errors
The rapid rise of artificial intelligence (AI) has inserted a new character into people's lives: the chatbot.
- How Gen AI is Disrupting B2B Buying Decisions
Organizations built around sales relationships and channel control must adapt to new types of recommendation systems.
- More AI Code, Less Certainty: Why Production Measurement Is the New Code Review
More AI Code, Less Certainty: Why Production Measurement Is the New Code Review DevOps.com
Score: 50🌐 MovesJun 12, 2026https://devops.com/webinars/more-ai-code-less-certainty-why-production-measurement-is-the-new-code-review/ - Amazingly, Apple may emerge unscathed from its AI mess
Hello again from Fast Company and welcome back to Plugged In . Throughout its 15-year history, Siri has often seemed to be cursed , or at least an eternal underachiever . Few products have had so much unfulfilled potential for so long. Yet none of the AI assistant’s previous disappointments rivaled the one that began two years ago, when Apple declared that Siri had entered a new era, showed off a profusion of new AI-powered capabilities . . . and then failed to ship them . But Apple is finally on the cusp of putting this epic vaporware fiasco behind it. Judging from the Siri update and other Apple Intelligence AI features unveiled in this week’s keynote at its WWDC developer conference, the company might even extricate itself from the situation with surprisingly minimal damage. To recap the story thus far: At WWDC 2024, Apple depicted Siri complying with fancy, free-form requests such as, “Add this photo to the email I drafted to Madiha and Josh,” “Show me my photos of Stacy in New York wearing her pink coat,” and “Search Fitness+ for new yoga classes.” But when that fall’s updates for iOS, iPadOS, and MacOS arrived, most of Siri’s new features were AWOL. In March 2025, Apple acknowledged that they were unexpectedly challenging to implement and said it expected to roll them out in “the coming year.” That turned out to mean “by the end of 2026.” The lengthy postponement provided plenty of time for additional shoes to drop. Last December, Apple’s artificial intelligence chief, John Giannandrea—whose bona fides as a technologist had never translated into much progress for Siri— announced his retirement . His departure was part of a reshuffling that included the arrival of Amar Subramanya, a Google and Microsoft alum, as VP of AI. A few months later, Mike Rockwell, the Apple Vision Pro godfather, was put in charge of Siri. And in January of this year, Apple and Google announced that the latter company’s Gemini AI model and cloud technology would be ingredients in future Apple AI experiences, including Siri. All of that led to WWDC 2026, where AI dominated the keynote once again. Most notably, Apple essentially reannounced the version of Siri that had fallen into limbo in 2024. It didn’t exactly undersell the update, now dubbed Siri AI: In the keynote video, Rockwell called it “profoundly more capable” than previous incarnations. But at every turn, the company took pains to prove that what it was showing was real. Indeed, this year’s prerecorded keynote had a cinema verité feel, complete with pregnant pauses as Siri AI chugged away before responding to requests. At subsequent WWDC briefings, executives did plenty of live demos of Siri and other AI updates, helping to instill confidence that they actually, you know, worked. This eagerness to let the new software be judged for what it is extended to Apple’s policy on reviews. In the past, the company discouraged journalists who installed the developer betas from publishing their impressions, preferring that they wait for later, more polished public betas. This year, it didn’t attempt to impose that moratorium. My own early experiences with the new AI features—which I’ve been trying on my iPhone 16 Pro and iPad Pro—have been largely positive. In my experiments, Siri AI effectively handled the kind of multistep requests Apple has been touting, including ones such as, “Text some photos of Aunt Liz to Marie” and “Find Mark Wilson’s Gmail address and add him as a contact.” The assistant even checked messages in my iPhone’s voicemail to help it answer questions, a reminder that an AI assistant embedded in an operating system can know us better than one that is not. It can also look at what’s on-screen to answer questions such as, “Based on this article, who’s investing in SpaceX?” All of Apple’s new AI is slickly integrated into iOS and iPadOS features such as Spotlight rather than walled off in a chatbot. That said, there is a new Siri app, a rough counterpart to ChatGPT, Claude, Copilot, and company. As you might expect of an Apple product, it’s a tad cautious: For instance, I couldn’t get it to generate images of specific humans at all (including Lincoln and Cleopatra, who probably wouldn’t have complained). Most people who are already smitten with another AI bot probably won’t be tempted to switch, especially if they’re paying for advanced features Siri lacks. But I could see the app-ified version of Siri being popular with folks who might otherwise be casual users of the free versions of competitors. Apple’s pledge of privacy by design is another point in Siri AI’s favor. We already know that people are having sensitive conversations with AI on topics such as their health , making the technology far more intimate than plugging keywords into search engines ever was. Apple hands off aspects of AI that can’t be processed on its devices to its Private Cloud Compute infrastructure, where it isn’t stored and can’t be read by anyone, including Apple. Meanwhile, Google , Meta , and OpenAI have every incentive to use AI queries on free services for ad targeting purposes, an inherently unnerving business model. The sudden spurt of progress reflected in Siri and other AI features is possible only because of a sweeping under-the-hood upgrade spanning all of Apple’s platforms. During the WWDC keynote video and a follow-up ” tech talk ,” the Apple software chief Craig Federighi acknowledged Gemini’s role in powering this update. However, as he explained it, the Gemini large language model supports Apple’s own models—known as Apple Foundation Models—rather than supplanting them. Instead of intermingling with Google’s own Gemini infrastructure or even resembling it, Apple’s new AI stack is isolated on its own Nvidia-based servers in Google data centers. (As used by Google’s products, Gemini runs on machines powered by Google’s custom TPU chips.) Given how all-in Google is on promoting Gemini as its AI brand, I understand Federighi’s desire to dispel any impression that Apple Intelligence amounts to relabeled Google technology. It’s a little as if someone licensed the rights to 5 of Colonel Sanders’s 11 secret herbs and spices, then used them in a chicken dish that wasn’t KFC. No matter how tasty it was, people would be confused. Still, it’s also easy to see why Apple turned to Google to expedite the process of building a new AI framework for the iPhone, iPad, Mac, and other devices. Siri’s very longevity helps explain why reimagining it for the post-ChatGPT age proved so extraordinarily difficult: By 2026 AI standards, it had the engine of a Model T. The new platform Federighi outlined is clearly much more modern—capable of enabling not just the AI features Apple announced at WWDC, but many more it hasn’t even thought of yet. Ultimately, Apple’s AI push is about making the technology essential to myriad features across its hardware platforms. Other makers of operating systems and devices are still figuring that out, too, leaving Apple less behind than it might have been if they’d made more progress in the past two years. Microsoft, for instance, keeps ushering in the age of the AI PC , and yet it still doesn’t feel like it’s really here. On the AI-everywhere front, Apple’s most formidable rival is its partner Google. Less than three weeks before WWDC, the company announced 100 things —its count, not mine—at its I/O conference. The buzziest debutante was Gemini Spark , a personal AI agent, with access to data stored in Google products such as Gmail and Drive, that can work more independently on users’ behalf than anything Apple previewed at WWDC. However, WWDC wasn’t a washout on the AI agent front. Among its most intriguing announcements were the new ability to vibe code Shortcuts and Safari extensions for use on iPhones, iPads, and Macs, giving relatively nontechnical people the ability to create new features for their devices. In the first developer beta, at least, these tools are a tad rudimentary, but also useful, fun, and agentic by nature. So is another Safari AI addition, Notify Me , which can autonomously monitor webpages for changes, such as an out-of-stock product becoming available again. They’re not Spark, but they have a spark of life not always visible in past iterations of Apple Intelligence. A few days with early developer betas can tell you only so much about Apple’s OS updates, which are likely to evolve further before shipping in the fall. And with AI’s relentless pace, the company could fall behind again if it doesn’t have stellar WWDCs in 2027, 2028, and beyond. Regardless, the mojo it showed off this week was more significant news than any single AI feature. Now it just needs to turn it into momentum. You’ve been reading Plugged In , Fast Company ’s weekly tech newsletter from me, global technology editor Harry McCracken. If a friend or colleague forwarded this edition to you—or if you’re reading it on fastcompany.com—you can check out previous issues and sign up to get it yourself every Friday morning. I love hearing from you: Ping me at hmccracken@fastcompany.com with your feedback and ideas for future newsletters. I’m also on Bluesky , Mastodon , and Threads , and you can follow Plugged In on Flipboard. More top tech stories from Fast Company Anthropic’s Claude Fable 5 plays it too safe on safety, developers say Developers are complaining on social media that Claude Fable 5’s safety system is blocking benign prompts, from résumé edits to shopping lists. Read More → Alamo Drafthouse CEO clears the air on that controversial new phone policy Alamo Drafthouse loyalists are upset over a move to a digital ordering system, but the company says it’s on track to have its best year for box office revenue since 2019. Read More → The real reason enterprise AI is stuck The AI industry is still building from metaphors. And metaphors do not industrialize. Read More → Formlabs’ new 3D printer is poised to reshape manufacturing The Fuse X1 is designed to bring large-format SLS printing to more engineering teams and production lines. Read More → The startup that tried to fix food waste—and got hit by a disinformation campaign Apeel was quickly growing until a wave of social media attacks spread by wellness influencers. Here’s how it survived. Read More → The backlash against AI, in 4 charts Americans are quickly turning against artificial intelligence. Read More →
- Investing Opportunities in AI and the Productivity Boom
Investing Opportunities in AI and the Productivity Boom Barron's
- Chico Schools to Install Controversial AI-Powered Cameras
Chico Unified School District is moving ahead with a 10-year contract to replace its surveillance cameras with new AI-powered ones, though some parents and students are concerned about privacy and data security.
Score: 50🌐 MovesJun 12, 2026https://www.govtech.com/education/k-12/chico-schools-to-install-controversial-ai-powered-cameras - 'Botsitting' is destroying productivity as workers spend nearly a full day each week making AI 'usable'
'Botsitting' is destroying productivity as workers spend nearly a full day each week making AI 'usable' IT Pro
- UAE Ministry of Higher Education drives AI-powered digital transformation to streamline student and university services
UAE Ministry of Higher Education drives AI-powered digital transformation to streamline student and university services Gulf News
- The illusion of safety: What happens when LLMs say the right things for wrong reasons
One of the most misleading moments in AI deployment is when the model sounds exactly as it should. It uses careful language. It gives balanced caveats. It avoids prohibited phrasing. It appears measured, compliant, and responsible. The tone feels safe enough for internal rollout and polished enough for senior stakeholders to relax. At that point, […] The post The illusion of safety: What happens when LLMs say the right things for wrong reasons appeared first on e27 .
Score: 50🌐 MovesJun 12, 2026https://e27.co/the-illusion-of-safety-what-happens-when-llms-say-the-right-things-for-wrong-reasons-20260606/ - Anthropic's Claude Fable 5 costs twice as much for 5.7 percent more performance
Claude Fable 5 tops the Artificial Analysis Intelligence Index with 64.9 points and sets records in five of ten benchmarks. But the gain over Opus 4.8 is just 5.7 percent at double the token price. Safety filters with fallback routing push costs even higher. The article Anthropic's Claude Fable 5 costs twice as much for 5.7 percent more performance appeared first on The Decoder .
Score: 50🌐 MovesJun 12, 2026https://the-decoder.com/anthropics-claude-fable-5-costs-twice-as-much-for-5-7-percent-more-performance/ - The flattening: How AI is collapsing the middle of the risk function
Last year, I drew the org chart of the risk team I would hire if I were building from zero. I drew four boxes. Fifteen years earlier, when I first joined a risk team inside an Indonesian bank, the chart I worked under had eleven boxes — analysts feeding into officers feeding into heads feeding […] The post The flattening: How AI is collapsing the middle of the risk function appeared first on e27 .
Score: 50🌐 MovesJun 12, 2026https://e27.co/the-flattening-how-ai-is-collapsing-the-middle-of-the-risk-function-20260605/ - EY GDS launches ey.ai Center for Reimagination in Bengaluru
EY GDS launched the ey.ai Center for Reimagination (CFR) designed to help organisations experience how emerging technologies and industry shifts will reshape industries, workforces and business decisions. The Center, which is spread over 40,000 square feet, is located at EY GDS, Bengaluru. The post EY GDS launches ey.ai Center for Reimagination in Bengaluru appeared first on Express Computer .
Score: 50🌐 MovesJun 12, 2026https://www.expresscomputer.in/news/ey-gds-launches-ey-ai-center-for-reimagination-in-bengaluru/135961/ - WWDC 2026: Apple’s AI Credibility Test
Although WWDC is a developers’ conference, this year’s keynote was primarily a statement of intent about Apple’s future and demonstration of its AI credibility. The most important announcement this year was the new Siri AI. Apple rebuilt it from the ground up, trying to make AI feel native, useful and invisible across devices people already […] The post WWDC 2026: Apple’s AI Credibility Test appeared first on IDC .
Score: 50🌐 MovesJun 12, 2026https://www.idc.com/resource-center/blog/wwdc-2026-apples-ai-credibility-test/ - AI will not make GBS firms obsolete, but more important: BCG report
Global Business Services will evolve from service-delivery centres into enterprise AI orchestration hubs, the report said
- The Tech Download: Mistral's Arthur Mensch on agentic AI, chips and enterprise adoption
CNBC's Arjun Kharpal sat down with the chief of one of Europe's leading AI companies.
Score: 50🌐 MovesJun 12, 2026https://www.cnbc.com/2026/06/12/tech-download-mistral-mensch-agentic-ai-chips.html - New open-source tool accelerates testing for trustworthy artificial intelligence
New open-source tool accelerates testing for trustworthy artificial intelligence EurekAlert!
- Wealth gap and AI squeeze young Koreans
Wealth gap and AI squeeze young Koreans 매일경제
- Taiwan Excellence Brings Trusted AI Innovation to ITS America 2026, Strengthening Taiwan–U.S. Mobility Partnerships
Taiwan Excellence Brings Trusted AI Innovation to ITS America 2026, Strengthening Taiwan–U.S. Mobility Partnerships Automotive News
- The gender data gap and the need for representation in AI
What is often overlooked is that mitigating AI gender bias starts with the data itself.
Score: 49🌐 MovesJun 12, 2026https://www.techradar.com/pro/the-gender-data-gap-and-the-need-for-representation-in-ai - Despite his new trillionaire status, Elon Musk says money ‘will stop being relevant’ in the future because of AI
Despite his new trillionaire status, Elon Musk says money ‘will stop being relevant’ in the future because of AI Fortune
Score: 49🌐 MovesJun 12, 2026https://fortune.com/2026/06/12/elon-musk-trillionaire-spacex-ipo-money-one-day-irrelevant-ai-robotics/ - olmo-eval: An evaluation workbench for the model development loop
olmo-eval is an open evaluation workbench that helps model developers add, run, and analyze benchmarks across changing LLM checkpoints, extending OLMES from final-score reproducibility into the day-to-day model development loop.
- Coram AI plans 20 Sunnyvale hires after $35M funding round
The firm raised $35 million and plans to expand hiring in Sunnyvale. CEO Ashesh Jain says AI is making workers more productive, not reducing the need for talent.
Score: 48💰 MoneyJun 12, 2026https://www.bizjournals.com/sanjose/news/2026/06/12/coram-ai-35m-funding-sunnyvale-hiring.html?ana=brss_6150 - AI-written books divide publishing world
AI-written books divide publishing world CBC
- Neuro A.I. Moon Shots & Fable 5 Backlash
Neuro A.I. Moon Shots & Fable 5 Backlash Puck
Score: 48🌐 MovesJun 12, 2026https://puck.news/newsletter_content/neuro-ai-moon-shots-fable-5-backlash/ - London Tech Week 2026: the AI billions, the US build-out, and a royal first
London Tech Week wrapped its 12th edition this week, after three days of main-stage programming at Olympia from 8 to 10 June, with fringe events running across the city through Friday. The festival drew more than 30,000 people from over 130 countries, with 600-plus speakers. One theme swallowed the rest: AI featured in roughly half […] This story continues at The Next Web
- Ever wonder why ChatGPT is free? The answer is far more calculated than you might think
Ever wonder why ChatGPT is free? The answer is far more calculated than you might think Tom's Guide
- Dietsmann Brings its Energy Maintenance and Robotics Expertise to African Energy Week (AEW) 2026
Dietsmann Brings its Energy Maintenance and Robotics Expertise to African Energy Week (AEW) 2026
- Mythos Begets Fable, Cursor's Composer 2.5, Agents Building Agents
Mythos Begets Fable, Cursor's Composer 2.5, Agents Building Agents
- Japan’s real AI problem is not technology. It is trust.
Japan’s real AI problem is not technology. It is trust. The Japan Times
Score: 47🌐 MovesJun 12, 2026https://www.japantimes.co.jp/commentary/2026/06/12/japan/japans-real-ai-problem/ - Building Human-AI Collaboration for the Future of Work
Building Human-AI Collaboration for the Future of Work Gartner
- AOL cofounder Steve Case on AI— major upside, real risk, and ‘probably a net negative’ for jobs
AOL cofounder Steve Case on AI— major upside, real risk, and ‘probably a net negative’ for jobs Fortune
- Why Systems Of Record Aren’t Enough In The Age Of AI
Better decisions generate better data, which feeds stronger intelligence, which produces better decisions.
- Microsoft president responds to students’ distrust for AI
Microsoft’s president, Brad Smith, has reacted to student discontent with AI, telling today’s graduates that there is still a place for human creativity. Students across the US have booed speakers who talked up AI at their graduation ceremonies in recent months, including Google’s former CEO Eric Schmidt , the CEO of a record label , and a real estate executive . Smith hasn’t ventured out onto a podium to share his views, but in a lengthy blog post, AI, Jobs and the Next Generation , acknowledged students’ concerns about their futures. He said that, just as painting survived the arrival of photography, so will the job market survive the arrival of AI. “While it may feel unfair that the job market is so uncertain, you were made for this moment. Technology is second nature to your generation. Constant change has taught you how to adapt quickly,” he wrote. He also used the blog to promote a book written by his colleagues Ryan Roslansky and Aneesh Raman on how to get ahead at work in the age of AI. The corporate world will see massive changes, he said: “This includes AI automation of tasks in current entry-level positions and, especially in the tech sector, corporate pressure to reduce headcount to help pay for AI’s enormous capital expenditures.” Some of those changes are already here. In the past six months, we have seen massive job losses at Oracle , at Meta and at AWS . There are no signs of any let-up: Last month saw the tech industry shed more than 38,000 jobs . Students contemplating their future will find little comfort in Smith’s optimistic words, particularly as his essay shows that Microsoft is not making any changes to its AI program going forward. This article first appeared on Computerworld.
Score: 46🌐 MovesJun 12, 2026https://www.cio.com/article/4184565/microsoft-president-responds-to-students-distrust-for-ai-2.html - Japanese parents turn to AI for clues hidden in babies’ cries
Japanese parents turn to AI for clues hidden in babies’ cries The Straits Times
Score: 46🌐 MovesJun 12, 2026https://www.straitstimes.com/asia/east-asia/japanese-parents-turn-to-ai-for-clues-hidden-in-babies-cries?ref - AI 교육받은 직원 74%, 비승인 툴 사용…아는 만큼 규칙도 어긴다
지난해 한 메신저 앱 기업의 엔지니어는 인증된 기술 업계 종사자들이 이용하는 익명 커뮤니티 블라인드에 글 을 올렸다. 챗GPT, 클로드, 제미나이 사용을 제한하는 기업이 모두 그런 것인지, 아니면 자신이 다니는 회사만 그런 것인지 묻는 내용이었다. 그가 근무하던 회사는 해당 도구 사용을 금지하는 대신 챗GPT 기반의 내부 대체 서비스를 제공했다. 그러나 이 엔지니어는 업무 속도가 느려진다는 이유로 이를 선호하지 않았다. 그는 “사실상 별 도움이 되지 않았다”고 회상했다. 해당 게시글에는 곧 다른 기업의 기술 인력이 대거 의견을 남겼다. 이들은 회사가 승인한 AI 도구가 지나치게 제한적이거나 유용한 기능 상당수가 제거돼 있다는 점에 불만을 나타냈다. 일주일 뒤 같은 엔지니어는 해결책을 들고 다시 게시판을 찾았다. 웹어셈블리(WebAssembly) 기반 대규모언어모델(LLM) 엔진을 활용해 브라우저 안에서 코딩 모델을 완전히 실행하는 방법을 찾아낸 것이다. 대화 내용은 로컬에 저장되고 외부 네트워크 트래픽도 발생하지 않아 회사가 이를 탐지할 수 없었다. 그는 게시글에 “즐거운 코딩 되길 바란다”며 “기능이 궁금하면 개인 메시지를 보내 달라”고 적었다. 생성형 AI의 기능을 가장 잘 이해하는 직원일수록 오히려 관련 규정을 우회하거나 위반할 가능성이 높다. 엔지니어는 물론, 어쩌면 의외로 의무 AI 교육을 이수한 직원 역시 공식적인 통제 장치를 엄격한 규칙이라기보다 업무 속도를 높이기 위해 넘어야 할 장애물로 인식하는 경우가 많다. 데이터 및 분석 기업 렉시스넥시스(LexisNexis) 보고서에 따르면 AI 교육을 받은 직원의 74%가 승인되지 않은 AI 도구를 사용한 경험이 있다고 답했다. 교육을 받지 않은 직원의 비율은 17%에 그쳤다. 넥시스 솔루션즈(Nexis Solutions) 제품 담당 부사장 다니 맥코믹은 “문제의 핵심은 직원의 역량과 기업 환경에 적합한 도구 사이의 격차”라며 “AI 도구에 대한 이해도가 높은 직원일수록 이를 실험하고 업무 흐름에 통합하려는 경향이 강하다”고 설명했다. AI 교육은 생성형 AI에 대한 초기 거부감이나 주저함을 줄이는 역할을 하는 것으로 보인다. 이러한 심리적 장벽은 AI 도입을 가로막는 요인이 될 수 있다. 맥코믹은 “교육이 위험을 초래한다는 의미는 아니다”라며 “많은 기업이 대응할 준비를 갖추기 전에 수요가 더 빠르게 드러난다는 점이 핵심”이라고 말했다. 이 같은 상황에서 CIO는 AI 활용을 장려하는 동시에 사용 방식을 통제해야 하는 어려운 과제에 직면해 있다. 직원들이 생성형 AI에 익숙해질수록 전면 금지와 같은 기존 접근 방식은 더 이상 효과를 발휘하지 못할 수 있으며, 오히려 역효과를 낼 가능성도 있다. 보다 생산적인 접근법은 섀도우 AI의 긍정적인 측면에 주목하는 것이다. P&G CIO 세스 코언(Seth Cohen)은 승인되지 않은 AI 도구 사용이 직원들이 해당 기술의 가치를 인식하고 더 빠르게 일하려는 의지를 보여주는 신호일 수 있다고 분석했다. 그는 “중요한 것은 그런 학습 경험을 기업에 적합하고 확장 가능한 시스템 안으로 흡수하는 것”이라고 설명했다. 하지만 실험을 장려하는 동시에 민감한 데이터를 보호하고, 점점 더 분산되는 AI 환경에 대한 통제력을 유지해야 하는 CIO들에게 그러한 시스템을 구축하는 일은 쉽지 않은 과제로 남아 있다. AI 교육 프로그램을 새롭게 설계하라 기업이 AI 활용 과정에서 직면하는 가장 큰 과제 중 하나는 조직 내 활용 수준의 편차가 크다는 점이다. 일부 팀은 이미 AI를 일상 업무 프로세스에 깊숙이 통합했지만, 다른 팀은 여전히 활용을 주저하거나 적극적으로 참여하지 않고 있다. 넥시스 솔루션즈(Nexis Solutions) 제품 담당 부사장 다니 맥코믹(Dani McCormick)은 “이러한 불균형이 승인되지 않은 AI 사용이 가장 두드러지게 나타나는 영역이며, 동시에 조직 차원의 정렬을 개선할 수 있는 가장 큰 기회가 존재하는 곳”이라고 설명했다. 이 격차를 해소하는 방법 중 하나는 AI 활용의 기술적·윤리적 측면을 모두 다루는 실습 중심의 AI 교육 프로그램이다. 이러한 프로그램은 직원들에게 승인된 AI 도구를 일상 업무에 통합하는 방법을 가르치는 동시에, 민감한 데이터 보호와 규제 준수, 조직 전반의 투명성과 책임성 확보를 위해 왜 승인된 플랫폼을 사용해야 하는지 설명해야 한다. P&G CIO 세스 코언(Seth Cohen)은 “교육은 직원이 의사결정을 개선하고 혁신을 가속화하며 실행력을 높이는 등 실제 업무에 적용할 수 있을 때 가장 효과적”이라고 말했다. 이러한 교육은 기술 인력에만 국한돼서는 안 된다. 생성형 AI 도구가 빠르게 대중화되면서 정식 기술 교육을 받지 않은 직원도 스스로 다양한 AI 서비스를 시험해 보는 사례가 늘고 있기 때문이다. 이는 많은 CIO가 공통적으로 관찰한 변화이기도 하다. 디트로이트시 CIO 아트 톰프슨(Art Thompson)은 “교육을 받지 않은 직원의 약 30%가 AI 기능에 관심을 보이며 직접 탐색하고 있는 것으로 보인다”고 말했다. 이어 “진정한 목표는 직원들이 기술을 책임감 있게 활용할 수 있도록 지원하는 것”이라며 “그렇지 않으면 섀도우 AI 생태계가 더욱 커지고 현재보다 가시성은 오히려 낮아질 것”이라고 경고했다. 효과적인 AI 교육 프로그램은 판단력, 거버넌스, 신뢰라는 세 가지 요소를 동시에 다뤄야 한다. 또한 직원들이 조직과 파트너, 그리고 AI 도구가 작동하는 더 넓은 생태계를 이해할 수 있도록 해야 한다. 직원들은 자신의 선택이 데이터 보안, 고객 신뢰, 규제 준수, 비즈니스 관계에 어떤 영향을 미칠 수 있는지 이해해야 한다. 톰프슨은 여전히 많은 직원이 AI 공급업체가 정보를 수집하는 방식이나 AI 결과물을 검증하는 방법을 제대로 이해하지 못하고 있다고 지적했다. 그는 “규칙을 마련하는 것은 좋은 출발점이지만, 직원들이 책임감 있게 도구를 사용하려면 그 지침의 의미를 이해해야 한다”며 “사업 부서가 거버넌스에 적극 참여하고 IT 문화의 일부가 되는 것이 이를 정착시키는 좋은 방법”이라고 설명했다. 여러 CIO는 규정 집행 역시 신중하게 이뤄져야 한다고 강조한다. AI 거버넌스·리스크·컴플라이언스(GRC) 플랫폼 기업 로직게이트(LogicGate)의 최고경영자(CEO) 겸 공동 설립자 맷 쿤켈(Matt Kunkel)은 “직원들이 AI 실험으로 인해 징계를 받을 수 있다고 우려한다면 사용을 중단하는 것이 아니라 단지 이를 숨길 뿐”이라며 “기업은 직원들이 불이익에 대한 두려움 없이 AI 활용 사실을 공개할 수 있는 환경을 조성하고, 잠재적인 AI 위험을 발견해 알리는 직원에게는 적절한 보상을 제공해야 한다”고 말했다. 직원이 원하는 AI 환경을 제공하라 더 나은 교육 프로그램과 강력한 거버넌스 체계를 구축하는 것만으로는 충분하지 않다. 기업은 직원들이 왜 섀도우 AI를 찾게 되는지 그 근본적인 원인도 해결해야 한다. 대부분의 경우 직원들은 업무를 더 빠르게 처리하고 일상 업무의 불편함을 줄여주는 도구를 원하기 때문에 승인되지 않은 AI 서비스를 사용한다. 많은 경우 직원이 개인 비용을 들여 AI 도구를 구독한다면, 이는 회사가 제공하는 공식 시스템만으로는 필요한 기능을 충족하지 못하고 있다는 의미일 수 있다. UKG CIO 프라카시 코타(Prakash Kota)는 “이는 위험 요소이면서 동시에 놓친 기회이기도 하다”며 “섀도우 AI는 직원이 하고자 하는 일과 조직이 실제로 가능하게 해주는 일 사이의 간극에서 성장한다”고 설명했다. 코타는 이러한 현상을 직원들이 무엇을 달성하려고 하는지, 그리고 공식 도구가 어떤 부분에서 부족한지를 더 잘 이해할 수 있는 기회로 받아들여야 한다고 강조했다. IT 서비스 기업 블루 맨티스(Blue Mantis) CIO 리처드 에이모스(Richard Amos)도 같은 견해를 보였다. 에이모스는 “우선 승인된 도구가 사용하기 어렵거나 기능이 제한적이거나, 제공 절차가 느린지 확인해 볼 것”이라며 “만약 그렇다면 직원들은 다른 대안을 찾게 된다. 사람은 본능적으로 업무를 더 효율적으로 처리할 방법을 찾기 마련”이라고 말했다. 또한 에이모스는 승인되지 않은 AI 도구를 사용하는 직원 대부분이 악의적인 의도를 가진 것은 아니라고 지적했다. 섀도우 AI는 규정에 대한 반발심보다는 호기심, 업무상의 불만, 또는 더 빠르게 일하고자 하는 욕구에서 비롯되는 경우가 많다는 설명이다. 그는 “사용 사례를 정확히 파악한 이후에는 AI 거버넌스 위원회에서 이를 검토하고 향후 도입 과제(backlog)로 고려할 수 있다”고 말했다. 직원들이 비공식적으로 사용하는 AI 도구에 주목하는 것은 CIO에게 또 다른 이점을 제공한다. 해당 도구가 더 큰 거버넌스 또는 보안 문제로 발전하기 전에 새로운 기술 트렌드를 조기에 포착할 수 있기 때문이다. 직원들의 실험 활동을 파악하고 있는 기업은 실제로 어떤 도구가 현업에서 유용하게 활용되는지 더 정확하게 이해할 수 있다. 보안 기업 에버론 솔루션즈(Everon Solutions) CIO 라이언 프리츠(Ryan Fritts)는 “새로운 AI 도구가 등장하는 즉시 이를 발견할 수도 있다”고 말했다. 더 빠르게 움직여야 한다 여러 기업은 직원들이 기존 시스템의 부족한 부분을 보완하기 위해 승인되지 않은 AI 도구를 사용하고 있다는 사실을 확인한 뒤 AI 플랫폼을 업그레이드했다. 그러나 단순히 승인된 도구를 제공하는 것만으로는 충분하지 않다. 이러한 플랫폼은 직원들의 요구 변화와 AI 기술 혁신의 빠른 속도에 맞춰 지속적으로 발전할 수 있을 만큼 유연하고 적응력이 있어야 한다. P&G CIO 세스 코언(Seth Cohen)은 “구식 모델을 사용하는 플랫폼 위에서 디지털 솔루션을 구축하는 데 열정을 느끼는 사람은 없을 것”이라고 말했다. 일부 기업은 직원들에게 더 큰 선택권을 부여하기 위해 최신 상용 기반 모델(foundation model) 가운데 원하는 모델을 선택할 수 있도록 지원하고 있다. 또 다른 기업들은 직원들이 안전하게 실험할 수 있는 보안 AI 환경을 제공한다. 코언은 “명확한 프레임워크 안에서 자유를 허용하고, 실험 과정에서 얻은 학습 내용을 수집해 플랫폼 전반을 개선하는 데 활용해야 한다”고 설명했다. 궁극적인 과제는 통제와 자율성 사이에서 적절한 균형을 찾는 것이다. 직원들이 AI를 업무에 효과적으로 통합할 수 있도록 충분한 자유를 제공하면서도 필요한 감독 체계를 유지해야 한다. CIO는 조직이 불필요한 위험에 노출되지 않도록 하면서도 팀이 지속적으로 학습할 수 있는 시스템과 문화를 구축해야 한다. 또한 기업은 새롭게 등장하는 AI 애플리케이션을 일일이 추적하는 끝없는 게임에 매달리기보다 데이터 관리에 집중하는 편이 더 큰 효과를 얻을 수 있다는 조언도 나온다. 에버론 솔루션즈(Everon Solutions) CIO 라이언 프리츠(Ryan Fritts)는 “실질적인 변화를 만들어내는 작업은 데이터 영역에 있다”며 “어떤 데이터가 어디에 저장돼 있는지 명확히 파악하고, 해당 데이터에 대한 정책을 실시간으로 적용할 수 있어야 한다”고 말했다. 이어 “데이터 관리 체계를 제대로 갖추면 섀도우 AI에 대한 과도한 우려와 불안은 대부분 자연스럽게 줄어들게 된다”고 설명했다. dl-ciokorea@foundryco.com
- Why Avataar Is Bullish About Cracking AI Video And Outdoing Global Giants
The race in AI video generation is currently being dominated by global giants like OpenAI, Google and Chinese startups flush…
Score: 45🌐 MovesJun 12, 2026https://inc42.com/features/why-avataar-is-bullish-about-cracking-ai-video-and-outdoing-global-giants/ - First Lady Melania Trump’s Remarkable Week Empowering Youth through AI Challenge and Fostering the Future Accounts
First Lady Melania Trump had a remarkable week focused on empowering children and youth through two major accomplishments: the first ever Presidential AI Challenge National Campion Awards Ceremony at the White House and the launch of Fostering the Future Accounts for youth in foster care. Presidential AI Challenge: On Tuesday, June 9, Mrs. Trump honored […] The post First Lady Melania Trump’s Remarkable Week Empowering Youth through AI Challenge and Fostering the Future Accounts appeared first on The White House .