AI is exerting an ever greater influence on our lives, which is leading to growing concern over whether we can trust it to act fairly and reliably. Ethical hackers, AI audits, and bias bounties could help us keep a lid on the potential harms, say researchers.
Theres increasing awareness of the dangers posed by our reliance on AI. These systems have a worrying knack for picking up and replicating the biases already present in our society, which can entrench the marginalization of certain groups.
The data-heavy nature of current deep learning systems also raises privacy concerns, both due to their encouragement of widespread surveillance and the possibility of data breaches. And the black box nature of many AI systems also makes it hard to assess whether theyre working correctly, which can have serious implications in certain domains.
Recognition of these issues has led to a rapidly expanding collection of AI ethics principles from companies, governments, and even supranational organizations designed to guide the developers of AI technology. But concrete proposals for how to make sure everyone lives up to these ideals are much rarer.
Now, a new paper in Science proposes some tangible steps that the industry could take to increase trust in AI technology. A failure to do so could lead to a tech-lash that severely hampers progress in the field, say the researchers.
Governments and the public need to be able to easily tell apart between the trustworthy, the snake-oil salesmen, and the clueless, lead author Shahar Avin, from Cambridge University, said in a press release. Once you can do that, there is a real incentive to be trustworthy. But while you cant tell them apart, there is a lot of pressure to cut corners.
The researchers borrow some tried and tested ideas from cybersecurity, which has grappled with the issue of getting people to trust software for decades. One popular approach is to use red teams of ethical hackers who attempt to find vulnerabilities in systems so that the designer can patch them before theyre released.
AI red teams already exist within large industry and government labs, the authors note, but they suggest that sharing experiences across organizations and domains could make this approach far more powerful and accessible to more AI developers.
Software companies also frequently offer bug bounties, which provide a financial reward if a hacker finds flaws in their systems and tells them about it privately so they can fix it. The authors suggest that AI developers should adopt similar practices, offering people rewards for finding out if their algorithms are biased or making incorrect decisions.
They point to a recent competition Twitter held that offered rewards to anyone who could find bias in their image-cropping algorithm as an early example of how this approach could work.
As cybersecurity attacks become more common, governments are increasingly mandating the reporting of data breaches and hacks. The authors suggest similar ideas could be applied to incidents where AI systems cause harm. While voluntary, anonymous sharingsuch as that enabled by the AI Incident Databaseis a useful starting point, they say this could become a regulatory requirement.
The world of finance also has some powerful tools for ensuring trust, most notably the idea of third-party audits. This involves granting an auditor access to restricted information so they can assess whether the owners public claims match their private records. Such an approach could be useful for AI developers who generally want to keep their data and algorithms secret.
Audits only work if the auditors can be trusted and there are meaningful consequences for a failure to pass them, though, say the authors. They are also only possible if developers follow common practices for documenting their development process and their systems makeup and activities.
At present, guidelines for how to do this in AI are lacking, but early work on ethical frameworks, model documentation, and continuous monitoring of AI systems is a useful starting place.
The AI industry is also already working on approaches that could boost trust in the technology. Efforts to improve the explainability and interpretability of AI models are already underway, but common standards and tests that measure compliance to those standards would be useful additions to this field.
Similarly, privacy-preserving machine learning, which aims to better protect the data used to train models, is a booming area of research. But theyre still rarely put into practice by industry, so the authors recommend more support for these efforts to boost adoption.
Whether companies can really be prodded into taking concerted action on this problem is unclear. Without regulators breathing down their necks, many will be unwilling to take on the onerous level of attention and investment that these approaches are likely to require. But the authors warn that the industry needs to recognize the importance of public trust and give it due weight.
Lives and livelihoods are ever more reliant on AI that is closed to scrutiny, and that is a recipe for a crisis of trust, co-author Haydn Belfield, from Cambridge University, said in the press release. Its time for the industry to move beyond well-meaning ethical principles and implement real-world mechanisms to address this.
Image Credit: markusspiske / 1000 images
Read the original here:
How Ethical Hackers Could Help Us Build Trust in AI - Singularity Hub
- Singularity Future Technology Ltd (SGLY) is up 43.55% Tuesday In Premarket Trading - InvestorsObserver - March 18th, 2024 [March 18th, 2024]
- This Week's Awesome Tech Stories From Around the Web (Through March 16) - Singularity Hub - March 18th, 2024 [March 18th, 2024]
- Palia reaches over 3m players in six months thanks to "invaluable" Switch partnership - GamesIndustry.biz - March 18th, 2024 [March 18th, 2024]
- Beyond the Singularity: Exploring the Fusion of AI and Art - Hong Kong Standard - March 18th, 2024 [March 18th, 2024]
- Your Comprehensive Guide to Telos Staking Success | by Pizza Singularity Sapphire | Feb, 2024 - Medium - February 20th, 2024 [February 20th, 2024]
- BTS' Kim Taehyung's 'Singularity' Performance Featured in Harvard Professor's Book - BNN Breaking - February 20th, 2024 [February 20th, 2024]
- Title: Understanding the Singularity: Unveiling the Future of Humanity - Medium - December 12th, 2023 [December 12th, 2023]
- What happens at the center of a black hole? - Astronomy Magazine - December 12th, 2023 [December 12th, 2023]
- IBM Is Planning to Build Its First Fault-Tolerant Quantum Computer by 2029 - Singularity Hub - December 12th, 2023 [December 12th, 2023]
- 22 Laws of Singularity And How You Can Apply Them To Live A Better Life - Medium - December 12th, 2023 [December 12th, 2023]
- Singularity: Here's When Humanity Will Reach It, New Data Shows - March 31st, 2023 [March 31st, 2023]
- sentinelOne expands singularity marketplace with new SOAR, insider threat, training, and prioritization integrations - ZAWYA - March 31st, 2023 [March 31st, 2023]
- Reaching the Singularity May be Humanitys Greatest and Last ... - March 4th, 2023 [March 4th, 2023]
- Singularity: Explain It to Me Like I'm 5-Years-Old - Futurism - March 4th, 2023 [March 4th, 2023]
- SINGULARITY FUTURE TECHNOLOGY LTD. : Notice of Delisting or Failure to Satisfy a Continued Listing Rule or Standard; Transfer of Listing, Change in... - March 4th, 2023 [March 4th, 2023]
- Apple co-founder Steve Wozniak on Artificial Intelligence: Not worried about The Singularity, well still be in control - MacDailyNews - February 12th, 2023 [February 12th, 2023]
- Cauchy principal value - Wikipedia - January 4th, 2023 [January 4th, 2023]
- Singularity Future Technology Ltd. (SGLY) Stockholder Notice: Robbins LLP Reminds Investors of the Class Action Against Singularity Future Technology... - December 14th, 2022 [December 14th, 2022]
- ROSEN, A LEADING LAW FIRM, Encourages Singularity Future Technology Ltd. f/k/a Sino-Global Shipping America Ltd. Investors to Secure Counsel Before... - December 12th, 2022 [December 12th, 2022]
- Singularity (mathematics) - Wikipedia - November 23rd, 2022 [November 23rd, 2022]
- GitHub: Where the world builds software GitHub - November 21st, 2022 [November 21st, 2022]
- Review: The Singularities, by John Banville - The New York Times - October 25th, 2022 [October 25th, 2022]
- We need to manage AI better as we are approaching the Creative Singularity - RedShark News - October 15th, 2022 [October 15th, 2022]
- This Week's Awesome Tech Stories From Around the Web (Through October 15) - Singularity Hub - October 15th, 2022 [October 15th, 2022]
- Ferguson: A second helping of QB play from Thanksgiving weekend - CFL.ca - October 15th, 2022 [October 15th, 2022]
- Kanimozhi slams Union govt over conducting CGL exams in only Hindi and English - The News Minute - October 15th, 2022 [October 15th, 2022]
- New Bayonetta 3 Trailer Reveals An In-Universe Singularity, And Lots Of Witches - Gameranx - October 13th, 2022 [October 13th, 2022]
- Singularity Future Technology Announces Receipt of Nasdaq Notice of Delisting and Intention to Request Hearing - Yahoo Finance - October 13th, 2022 [October 13th, 2022]
- Six Recent Discoveries That Have Changed How We Think About Human Origins - Singularity Hub - October 13th, 2022 [October 13th, 2022]
- This AI Uses a Scan of Your Retina to Predict Your Risk of Heart Disease - Singularity Hub - October 13th, 2022 [October 13th, 2022]
- The Moon May Have Formed Just Hours After Earth Collided With a Protoplanet - Singularity Hub - October 13th, 2022 [October 13th, 2022]
- UK HealthCare hosting 2 nationally esteemed guests for 60th anniversary symposium - UKNow - October 13th, 2022 [October 13th, 2022]
- Why Transcend Fund believes the opportunity for game investments is only getting bigger - VentureBeat - October 13th, 2022 [October 13th, 2022]
- Austin Powers References in This Fool, Ramy, Bros - Vulture - October 13th, 2022 [October 13th, 2022]
- Bayonetta 3 Everything You Need to Know About this Bewitching Beat em Up - Wccftech - October 8th, 2022 [October 8th, 2022]
- Longtermism: The Future Is VastWhat Does This Mean for Our Own Life? - Singularity Hub - October 8th, 2022 [October 8th, 2022]
- Ohio creates elections integrity office while voter fraud is already 'exceedingly rare' - ideastream - October 8th, 2022 [October 8th, 2022]
- Elon Musk Warns of World War III - TheStreet - October 8th, 2022 [October 8th, 2022]
- Durga reminds us of our collective obligations - Daily Pioneer - October 8th, 2022 [October 8th, 2022]
- Glitch in the algorithm - The Bucknellian - October 8th, 2022 [October 8th, 2022]
- As the Robot Fry Cook Takes Over the Kitchen - Walter Bradley Center for Natural and Artificial Intelligence - October 8th, 2022 [October 8th, 2022]
- The Singularity Image Format (SIF) Selected as a Finalist in the HPCwire Readers' Choice Awards - PR Web - October 6th, 2022 [October 6th, 2022]
- The Singularity Image Format (SIF) Selected as a Finalist in the HPCwire Readers' Choice Awards - Benzinga - October 6th, 2022 [October 6th, 2022]
- The Singularity of the Dual Mandate - Federal Reserve Bank of San Francisco - October 6th, 2022 [October 6th, 2022]
- What Is the Quirk Singularity Doomsday Theory in My Hero Academia? Explained - Twinfinite - October 6th, 2022 [October 6th, 2022]
- Introduction to week 4 of Tekedia Mini-MBA: Exponential Technologies and Singularity - Tekedia - October 6th, 2022 [October 6th, 2022]
- A $500 Million International Project Will Create the Most Detailed Map of the Brain Ever - Singularity Hub - October 6th, 2022 [October 6th, 2022]
- There Are Cheaper, More Sustainable Ways Than Desalination to Meet Our Water Needs - Singularity Hub - October 6th, 2022 [October 6th, 2022]
- #3DStartup: Unlimited Tomorrow and its 3D Printed Bionic Prosthetic Arm - 3Dnatives - October 6th, 2022 [October 6th, 2022]
- First As Parody, Then As Free Speech: The Onion Goes To The Supreme Court. It's About As Awesome As You'd Suspect. - Above the Law - October 6th, 2022 [October 6th, 2022]
- 10 Hot Cyber Threat Intelligence Tools And Services In 2022 - CRN - October 6th, 2022 [October 6th, 2022]
- Nintendo comes out strong with its offering of upcoming titles - The UML Connector - October 6th, 2022 [October 6th, 2022]
- The metaverses evolutionary roots could aid its success | Mint - Mint - October 6th, 2022 [October 6th, 2022]
- The Tech That Will Push VR to the Limits of the Human Eye - Singularity Hub - September 27th, 2022 [September 27th, 2022]
- Super-Earths Are Bigger and More Habitable Than Earth, and Astronomers Are Discovering More of the Billions They Think Are Out There - Singularity Hub - September 27th, 2022 [September 27th, 2022]
- Scientists Have Long Dreamed of a Memory Prosthesis. The First Human Trials Look Promising - Singularity Hub - September 27th, 2022 [September 27th, 2022]
- NASA's DART Spacecraft Will Smack an Asteroid at 14,000 MPH Todayand You Can Watch - Singularity Hub - September 27th, 2022 [September 27th, 2022]
- Why You Should Read This: 'The Gold Coast' - Alta Magazine - September 27th, 2022 [September 27th, 2022]
- The Multitude of Stromae - Mail and Guardian - September 27th, 2022 [September 27th, 2022]
- Warhammer 40,000: Space Marine is Getting a Location-based VR Experience Next Year - Road to VR - September 27th, 2022 [September 27th, 2022]
- What the pandemic taught us about teaching (On the Other Hand) - Montclair Local - September 27th, 2022 [September 27th, 2022]
- Where to get the Boom Sniper in Fortnite Chapter 3 Season 4 - The Nerd Stash - September 27th, 2022 [September 27th, 2022]
- Machine learning has predicted the winners of the Worlds - CyclingTips - September 27th, 2022 [September 27th, 2022]
- Vaonis Vespera Review: Easy To Use But with Underwhelming Results - PetaPixel - September 27th, 2022 [September 27th, 2022]
- The biggest problem with gravity and quantum physics - Big Think - September 27th, 2022 [September 27th, 2022]
- V gets inked on his butt cheek? BTS star spills the details about his friendship tattoo - Zoom TV - September 27th, 2022 [September 27th, 2022]
- Robots & Humans: Are we heading towards Singularity? - INDIAai - September 20th, 2022 [September 20th, 2022]
- This Week's Awesome Tech Stories From Around the Web (Through September 17) - Singularity Hub - September 20th, 2022 [September 20th, 2022]
- A Swedish Company Wants to Transform Offshore Wind With Vertical-Axis Turbines - Singularity Hub - September 20th, 2022 [September 20th, 2022]
- Decarbonizing the Energy Sector by 2050 Could Save the World $12 Trillion - Singularity Hub - September 20th, 2022 [September 20th, 2022]
- This Sleek Solar Car Goes 600 Miles on a Charge and Is Gearing Up for Production - Singularity Hub - September 20th, 2022 [September 20th, 2022]
- Meta Built an AI That Can Guess the Words You're Hearing by Decoding Your Brainwaves - Singularity Hub - September 20th, 2022 [September 20th, 2022]
- Gory Throwback FPS 'Prodeus' Exits Early Access This Thursday, Launching for PC and Consoles [Trailer] - Bloody Disgusting - September 20th, 2022 [September 20th, 2022]
- Meet the New Vespera Telescope From Vaonis - Universe Today - September 20th, 2022 [September 20th, 2022]
- Of God and Machines - The Atlantic - September 20th, 2022 [September 20th, 2022]
- Humans Destroyed Forests for Thousands of Years. We Can Become the First Generation to Expand Them - Singularity Hub - September 20th, 2022 [September 20th, 2022]
- Intel Core i9-13900K beats the 12900K by +10% in CPU-bound games - KitGuru - September 20th, 2022 [September 20th, 2022]
- INVESTIGATION ALERT: The Schall Law Firm Encourages Investors in Singularity Future Technology Ltd. with Losses of $100000 to Contact the Firm -... - August 15th, 2022 [August 15th, 2022]
- How To Open Singularity Rock in Tower of Fantasy - Attack of the Fanboy - August 15th, 2022 [August 15th, 2022]
- Intel i9-13900K Tested in Ashes Of The Singularity, Offers Neglible Increments in Performance - Appuals - August 15th, 2022 [August 15th, 2022]