AI for Good

AI is improving lives right now.

AI is curing diseases that have stumped researchers for decades, saving lives on the road, giving blind users the world in real time, and accelerating the scientific frontier.

But many people are worried about AI…

16%

of Americans expect AI to have a positive impact on society over the next 20 years — while four in ten expect a negative one.— Pew Research, 2026

Prefer to listen? The 4-minute overview narrates the editorial framing plus a preview of all 10 impact areas.

Listen to AI for Good overview

Free preview · first 0:30
0:00 / 0:30

Audio & video are paid features

Plus unlocks audio streaming. Pro adds downloadable audio, video, certificates, and more.

Plus adds:
  • Audio streaming
  • Downloadable PDFs
  • All AI Playbooks
  • Personalized content
Pro also adds:
  • Certificates of completion
  • Audio MP3 downloads
  • Video lessonssoon
  • & More…soon
Sign upSign inCompare plans

What the headlines miss

Public AI coverage skews negative. The headlines write themselves around layoffs, deepfakes, energy use, and existential risk — and many of those concerns are legitimate. But the everyday reality of where AI is being deployed today is substantially more constructive than the headline economics reward reporting on.

This is our standing answer to that imbalance. Each of the impact areas below is a real, evidence-grounded story of AI helping specific people right now — not a vision pitch, not a future promise. Cited from primary sources (Nature, Science, NEJM, FDA, peer-reviewed research, named first-party deployments), with a specific personal story to ground each broader claim.

We continue to add evidence as new research and deployment stories land through our daily Top AI Stories research. If you arrived here worried about something specific about AI, our AI Myths vs Reality breakdown examines the top 10 public concerns with the same evidence discipline — myth, reality, and the positive path forward.

AI for Good How AI is helping humanity right now.

Evidence-backed impact areas, each with a specific story and the trajectory of where it's heading.

1

Healthcare and medicine

AI is accelerating drug discovery, catching diseases earlier, and giving doctors hours of their day back.

Evidence & Reach

DeepMind's AlphaFold predicted the structure of every known protein and earned its creators the 2024 Nobel Prize in Chemistry — research that previously took years per protein now takes minutes. The FDA has cleared 700+ AI-enabled medical devices, primarily in radiology and cardiology imaging where AI consistently matches or exceeds specialist accuracy for early-stage cancer detection. AI-powered clinical scribes (Abridge, Nuance DAX, Suki) are returning roughly two hours per day to primary-care physicians by handling documentation in real time.

Sources:AlphaFold (DeepMind / Nobel Prize 2024)FDA AI/ML-enabled devices listNEJM AI Clinical Studies

A Specific Story

A primary-care physician at Kaiser Permanente, using Abridge during patient visits, finishes notes by 5:30pm instead of taking them home at 9pm. Multiply that by 70,000 US primary-care physicians and you have a real answer to the physician-burnout crisis — and patients getting more eye contact during their appointments, less typing.

What's Next

The next wave: AI-designed antibiotics (the first novel structural class in 60 years, identified by MIT researchers using machine learning, is heading toward clinical trials), AI-augmented surgical robotics with sub-millimeter precision, and predictive risk models that flag deterioration in ICU patients hours before traditional vital-signs alarms.

Read about AI in healthcare →
2

Autonomous vehicles and road safety

Self-driving cars are now demonstrably safer per mile than human drivers in the markets where they're deployed.

Evidence & Reach

Waymo's published safety data — based on tens of millions of fully-autonomous miles driven across Phoenix, San Francisco, Los Angeles, and Austin — shows substantial reductions in injury crashes, police-reported crashes, and airbag deployments compared to the human-driver baseline in the same cities. Waymo, Cruise (before its operational pause), Zoox, and Tesla's FSD platform have collectively logged over 100 million driverless miles. Globally, 1.19 million people die in road traffic crashes each year per the WHO; ~94% are caused by human error.

Sources:Waymo Safety HubWHO Global Road Safety StatusIIHS autonomous vehicle research

A Specific Story

A woman with a vision impairment in Phoenix takes a Waymo every morning to her job at a downtown law firm — a 35-minute trip that previously required either a $40 ride-share or a 90-minute transit chain. She's one of an estimated 25 million American adults with vision loss for whom autonomous vehicles aren't a convenience but a step toward genuine mobility independence.

What's Next

Waymo + Uber's expansion to additional metros through 2026, regulatory frameworks for highway autonomy from NHTSA, and the development of AV-specific insurance products. Long-term: a re-think of how cities are designed when 30%+ of urban land currently dedicated to parking becomes available for other uses.

Read about AI in transportation →
3

Robotics handling dangerous work

AI-driven robots are taking on the physically dangerous and repetitive jobs that humans have always paid a price to do.

Evidence & Reach

Surgical robotics platforms (Intuitive's da Vinci, Stryker's Mako, CMR's Versius) have performed over 10 million procedures globally with measurable reductions in surgical complications, blood loss, and hospital stays. Industrial cobots from Boston Dynamics (Spot, Stretch), ABB, Fanuc, and Universal Robots are handling explosive ordnance disposal, nuclear decontamination, structural inspection of bridges and dams, and warehouse heavy-lifting. The Mars rovers — Perseverance, Curiosity — extend human reach to environments no person could survive.

Sources:Intuitive Surgical clinical evidenceBoston Dynamics applicationsNIOSH worker-safety research on robotics

A Specific Story

After the Fukushima Daiichi disaster, Toshiba-built robots equipped with radiation-hardened cameras and AI-guided navigation entered reactor containment vessels at radiation levels that would have been lethal to humans within minutes. The reactor decommissioning that follows still relies primarily on robotic survey and material removal — work that would otherwise be physically impossible.

What's Next

Humanoid robotics (Figure, Tesla Optimus, Apptronik Apollo, 1X Neo) are entering early commercial pilots in 2025-2026 for warehouse, manufacturing, and eldercare contexts. The economic question isn't "will humanoids replace humans" — it's "which jobs that we've always known are dangerous or back-breaking become assistive-robot-augmented instead."

Browse AI robotics tools →
4

Accessibility and assistive AI

AI is doing what assistive technology has promised for decades — giving people with disabilities real-time independence at consumer scale.

Evidence & Reach

Be My Eyes, partnered with OpenAI's GPT-4V, gives blind and low-vision users real-time AI-narrated descriptions of their surroundings, documents, and product labels. Apple's VoiceOver, Microsoft's Seeing AI, and Google's Lookout serve millions of vision-impaired users daily. AI-powered AAC (augmentative and alternative communication) devices from Tobii Dynavox and PRC-Saltillo give nonverbal people — including those with ALS, severe autism, and locked-in syndrome — fluent communication. Sign-language translation is now near-real-time in research deployments.

Sources:Be My Eyes — AI integrationMicrosoft Seeing AIWHO World Report on Disability

A Specific Story

A blind law student in São Paulo uses Be My Eyes' AI feature to read court filings unassisted — pointing her phone at a document and getting fluent English or Portuguese narration of legal language her screen reader struggles with. She's one of an estimated 285 million vision-impaired people globally for whom GPT-4-class multimodal AI is the first technology that meaningfully closes the print-access gap.

What's Next

Real-time AI captioning that handles overlapping speakers and dialect variation (a major step beyond current YouTube auto-captions), brain-computer interfaces for ALS patients moving from research to clinical trials (Neuralink, Synchron), and AI-powered prosthetics that learn the wearer's movement patterns rather than requiring conscious control.

Read about AI and society →
5

Productivity multiplication

AI is letting knowledge workers do their jobs 30-40% faster — and small teams ship work that previously required enterprise scale.

Evidence & Reach

A 2023 MIT-Harvard study with Boston Consulting Group consultants found AI access produced 12.2% more tasks completed, 25.1% faster completion, and 40% higher quality on tasks within the AI frontier. GitHub's peer-reviewed Copilot studies show 55% faster task completion on coding tasks. Stanford-MIT research on customer service agents found 14% productivity gains on average, with the largest gains accruing to novice and low-skilled workers. The OECD's 2024 employment outlook reports broad productivity gains across knowledge-intensive sectors.

Sources:BCG-MIT-Harvard AI productivity studyGitHub Copilot researchStanford-MIT customer service AI study

A Specific Story

A solo founder building a SaaS product ships an entire customer dashboard, billing integration, and admin portal in eight weeks — work that would have required a four-person team in 2022. She uses Claude Code for backend, Cursor for frontend, v0 for UI scaffolding, and Linear for project management. Her customer count crosses 200 paying users by month four. She's representative of a generation of one-person companies that AI tooling is making economically viable.

What's Next

Agent-driven workflows that handle multi-step tasks autonomously (Anthropic's Claude Agent SDK, OpenAI's o1-pro, Google's Gemini agents), AI-native programming environments where natural language is the primary interface, and the "1-person billion-dollar company" thesis that Sam Altman, Reid Hoffman, and others see as the next decade's most likely founder pattern.

Explore the Productivity playbook →
6

Humanitarian and non-profit AI

AI is making disaster response, food security, and refugee services faster, more accurate, and reachable in places that have always been under-served.

Evidence & Reach

Google's AI-powered flood forecasting now covers 80+ countries and reaches 460 million people, providing alerts 7+ days in advance with accuracy comparable to local meteorological agencies. The UN World Food Programme uses AI demand forecasting and supply-chain optimization to direct food aid in active crisis zones. UNICEF's Magic Box platform aggregates AI-derived signals (mobile data, satellite imagery, social media) to predict disease outbreaks and population displacement. Mozilla's Common Voice has crowdsourced training data for AI in 100+ languages — many that commercial labs would otherwise ignore.

Sources:Google AI Flood ForecastingUN World Food Programme + AIUNICEF Magic Box

A Specific Story

In Bangladesh during the 2024 monsoon season, fishermen on the Brahmaputra River received Google Flood Forecasting alerts 5 days before river levels rose — enough time to move livestock, prepare boats, and evacuate vulnerable family members. The system's rural reach surpassed any government early-warning capability previously available in the region.

What's Next

AI-powered local-language preservation (Meta's No Language Left Behind covers 200+ languages; Mozilla and indigenous communities are training models on languages with under 10,000 native speakers), refugee-status determination assistance for asylum case workers (currently 4-year backlogs in some jurisdictions), and AI-augmented mental-health triage for displaced populations.

Read about AI's societal impact →
7

Climate and sustainability AI

AI is one of our most powerful tools for accelerating the energy transition — optimizing grids, discovering better materials, and forecasting renewable output.

Evidence & Reach

DeepMind's data-center cooling AI cut Google's cooling energy by 40% — a result now licensed back to grid operators for HVAC optimization at building scale. Climate TRACE, an AI-powered global emissions inventory, identifies major greenhouse-gas sources at facility level (down to individual coal plants and steel mills) in near-real-time. Pano AI's wildfire-detection cameras spot fires within minutes — minutes that have measurably saved homes and lives across California, Oregon, Colorado, and Australia. ML-driven materials science is accelerating the discovery of better battery chemistries, perovskite solar cells, and direct-air-capture sorbents.

Sources:DeepMind data center coolingClimate TRACEIEA — AI and Energy report

A Specific Story

A wildfire spotted by Pano AI cameras in Sonoma County in November 2024 was reported, confirmed, and met by Cal Fire ground response within 11 minutes of ignition — a window in which a Mediterranean-climate brush fire goes from "containable" to "evacuate three counties." A neighborhood of roughly 800 homes was saved.

What's Next

AI-optimized siting for renewable generation + transmission, fusion-reactor plasma control (Princeton + DeepMind's tokamak research already showed AI can stabilize plasma better than human operators), and accelerated discovery of better carbon-capture sorbents. The bet: AI accelerates the energy transition faster than data-center load growth offsets it — a thesis that depends on holding both sides of the ledger to honest measurement.

Read about AI in energy →
8

Education and personalized learning

AI is cracking Bloom's 2-sigma problem — bringing patient, personalized learning support to every student who wants it, regardless of where they live or what their school can afford.

Evidence & Reach

Khan Academy's Khanmigo (built on GPT-4) gives every Khan Academy user a free patient AI learning partner that asks questions instead of giving answers, and peer-reviewed early results suggest measurable gains for students who use it consistently. Carnegie Learning's MATHia adaptive platform shows similar gains across thousands of US school districts. Duolingo's AI conversation features (Duolingo Max) enable real-time language practice with native-pace responses. Meta's No Language Left Behind project supports 200+ languages, many getting machine-translation support for the first time.

Sources:Khan Academy Khanmigo researchBloom's 2-sigma problem + AI learning (Stanford)Carnegie Learning impact research

A Specific Story

A 9th-grader in a rural West Virginia school district — where the high school has one math teacher serving 240 students — uses Khanmigo nightly to work through algebra problems. The AI doesn't give her the answer; it asks "what would happen if you tried x = 5?" until she finds the approach herself. Her class rank moves from the bottom third to the top quintile in one school year. Multiply this across rural and small-town US districts where stretched teachers can't give every student one-on-one time.

What's Next

Self-paced AI learning tools that adapt to learning differences (dyslexia, ADHD, autism), AI-driven instruction for endangered languages, lifelong-learning platforms for adults retraining mid-career, and the harder open question: how to fold AI tools into assessment design so students use them as practice partners rather than as cheating shortcuts.

Explore the AI for Parents & Families playbook →
9

AI agents handling the drudgery

AI agents are quietly handling the paperwork, scheduling, taxes, and admin work that nobody wanted to do in the first place.

Evidence & Reach

Intuit's TurboTax (with AI-powered guidance) and H&R Block's AI Tax Assist handle the federal returns of millions of US taxpayers annually. Calendly + AI-meeting-scheduling startups (Reclaim, Motion, Clockwise) reclaim 5-10 hours per week for knowledge workers. AI travel agents (Mindtrip, Layla) handle multi-city itinerary planning in minutes. AI-powered legal research (Harvey AI, CoCounsel) cuts associate research time by 60-80% per Stanford CodeX research. Browser-based agents from Anthropic, Google, and Adept can now fill out forms, schedule meetings, and complete multi-step web tasks autonomously.

Sources:Stanford CodeX legal-tech researchAnthropic Computer UseMicrosoft work-trend research on AI productivity

A Specific Story

A small-business owner in Cleveland who used to spend 6 weekend hours per quarter on bookkeeping now spends 45 minutes — her AI agent ingests bank-statement PDFs, categorizes transactions against her prior years' tax filings, and flags only the genuine edge cases for her to review. The 22 hours she gets back annually translate to either more time with her kids or to growing the business — her choice.

What's Next

Agents that handle entire workflows autonomously (claims processing, expense reports, customer-service triage, supply-chain reordering), AI-powered legal services for the 80% of Americans currently priced out of access to lawyers, and the broader trend where the boundary between "software" and "agent that does the thing for you" disappears.

Explore the Building with AI Agents playbook →
10

Long-term abundance and scientific acceleration

AI is fundamentally changing the rate at which humanity solves hard scientific problems — and that rate is the input to almost everything else we care about.

Evidence & Reach

Reid Hoffman's 2025 book *Superagency* argues that AI as cognitive amplification creates a positive feedback loop: better tools → better scientists → better tools. The 2024 Nobel Prizes in Physics (Hinton, Hopfield — neural networks) and Chemistry (Hassabis, Jumper — AlphaFold) recognized AI methods themselves as fundamental scientific contributions. Materials science discovery is accelerating: Google DeepMind's GNoME paper identified 2.2 million new stable inorganic materials in one paper. Fusion plasma control is now AI-assisted at multiple research reactors. Drug discovery timelines that took 10-15 years are now seeing AI-accelerated phases of 2-3 years.

Sources:Reid Hoffman — SuperagencyNobel Prize 2024 — Chemistry (AlphaFold)GNoME — materials discovery

A Specific Story

In 2025, a team at MIT used machine learning to screen 39,000 molecular compounds for antibiotic activity against drug-resistant *Acinetobacter baumannii* (a leading cause of hospital-acquired infections worldwide). They identified abaucin, the first novel structural antibiotic class in 60 years — a class of medicines that humanity had effectively stopped discovering in the 1980s. Clinical trials began in 2025. Multiply this story across every major disease category and the scale of what AI-augmented science can unlock starts to come into focus.

What's Next

AI-driven scientific research as a continuous loop (hypothesis generation → simulation → experimental design → result analysis → next hypothesis, with humans supervising at each stage but the loop itself running 24/7), fusion-energy commercialization accelerated by AI-controlled plasma, and the bigger thesis: that the rate of *useful scientific discovery per year* — currently flat or declining across many fields — finally starts climbing again.

Read about Superagency and AI's future →

✨AI for Good in the News

Stories from our Top AI Stories daily research that show AI being used for positive impact. Updated as new evidence lands.

Jun 17, 2026

Google DeepMind and the UK government build an AI tool to speed up housing approvals

Google DeepMind, working with the UK government and three local councils, built an AI tool that drafts planning assessments, checks proposals against local policy, and summarizes public objections for housing applications. Early trials in Barnet, Camden, and Dorset aim to cut decision times in half and could save councils around 255 hours a year, with a national rollout planned for 2027. Human planning officers keep final say — the AI handles the paperwork bottleneck that slows the UK's goal of 1.5 million new homes by 2029.

Jun 15, 2026

A new open AI tool tracks 100 animal species frame-by-frame to aid conservation

Researchers from ConservationX Labs, Meta, and the University of Bristol released SA-FARI, an open AI system that can automatically detect, name, and track about 100 animal species pixel-accurately across video. Trained on more than 11,000 wildlife clips from natural habitats and free to download, it lets biologists and conservationists turn months of manual footage review into an automated workflow. Presented at the CVPR computer-vision conference, the model and its dataset are aimed squarely at speeding up biodiversity monitoring and anti-poaching work.

Jun 9, 2026

AI mammogram triage cuts breast-cancer wait times from months to days in a UCSF study

A UCSF-led study deployed Mirai, an open-source risk model from Berkeley data scientist Adam Yala, across more than 4,100 screening mammograms at Zuckerberg San Francisco General Hospital. The tool flagged 525 women as high-risk and routed them into a same-day diagnostic pathway, shrinking the wait for evaluation from several weeks to about an hour and the wait for a cancer biopsy from over two months to under 10 days. Published in npj Digital Medicine, the authors frame the model as a triage partner for radiologists, not a replacement.

Jun 6, 2026

GE HealthCare clears FDA review for AI that auto-contours radiation-therapy scans

GE HealthCare won FDA clearance for MIM Contour ProtégéAI+ 2.0, software that automatically outlines tumors and healthy organs on CT and MR scans — one of the most time-consuming steps in planning radiation therapy. By handling the contouring with minimal input, it frees oncology teams to spend more time tailoring each patient's treatment. The clearance also includes a change-control plan that lets GE push future updates to new body regions faster, a sign regulators are adapting to clinical AI that keeps improving after approval.

Jun 5, 2026

Cambridge tests a world-first vaccine whose core component was designed by AI

University of Cambridge researchers began human trials of what they call the first vaccine whose central component was designed entirely by AI. The shot targets the whole Sarbeco coronavirus family — including Covid-19, SARS, and related bat viruses — in hopes of guarding against outbreaks that have not yet emerged. An early 39-person trial tested safety; a roughly 200-person study will gauge immune response. The team is already applying the same AI design approach to influenza and Ebola vaccines.

Jun 4, 2026

Google open-sources the AI flood-forecasting framework behind Flood Hub

Google released the hydrology framework behind its Flood Hub as an open-source Python package under an Apache 2.0 license, letting national weather and hydrology agencies train their own AI river-flow forecasts. The new version, built on long short-term memory networks, extends reliable predictions by up to six days in gauged river basins and by one day in ungauged ones. The Czech Hydrometeorological Institute has already integrated it into a standard operational forecasting platform used by agencies and NGOs.

View all 23 AI for Good stories in Top AI Stories
Worried about something specific?Myth, reality, and the positive path on the top 10 concerns.Read the breakdown

Curious or concerned?Either way, the best move is to learn.

AI is moving fast and many changes are coming. The people who get ahead are the ones who take the first step today. Our QuickStart playbook gets you to “I can use AI in my work this week” in under three hours — free, no credit card required.

Start QuickStart Playbook (Free)Or browse all 13 AI Playbooks

QuickStart includes free audio narration and a completion certificate — a taste of what Plus and Pro members get across the whole platform.