AI Incident Roundup – December ‘22

Welcome to this month’s edition of The Monthly Roundup, a newsletter designed to give you a digestible recap on the latest incidents and reports of the AI Incident Database.

Estimated reading time: 5 minutes

🗞️ New Incidents

Emerging incidents that occurred last month:

Incident #416: Facebook's Job Ad Algorithm Allegedly Biased against Older and Female Workers

• What happened? Facebook's algorithm was alleged in a complaint by Real Women in Trucking to have selectively shown job advertisements disproportionately against older and female workers in favor of younger men for blue-collar positions.

Incident #419: Facebook's Automated Moderation Allowed Ads Threatening Election Workers to be Posted

• What happened? Facebook's automated moderating system failed to flag and allowed ads containing explicit violent language against election workers to be published.

Incident #430: Lawyers Denied Entry to Performance Venue by Facial Recognition

• What happened? Lawyers were barred from entry to Madison Square Garden after a facial recognition system matched them as employed by a law firm currently engaged in litigation with the venue.

Incident #432: Southwest Airlines Crew Scheduling Solver Degenerates Flight Network

• What happened? Southwest Airlines left passengers stranded for days throughout the flight network when Southwest crew scheduling software developed by General Electric repeatedly failed to recover from weather-induced flight cancellations.

📎 New Developments

Older incidents that have new reports or updates.

Original incident	New report(s)
Incident #279: TikTok’s “For You” Algorithm Exposed Young Users to Pro-Eating Disorder Content	TikTok self-harm study results ‘every parent’s nightmare’ – theguardian.com, 2022-12-15
Incident #413Thousands of Incorrect ChatGPT-Produced Answers Posted on Stack Overflow	Temporary policy: ChatGPT is banned – meta.stackoverflow.com, 2022-12-05 AI-generated answers temporarily banned on coding Q&A site Stack Overflow – theverge.com, 2022-12-05
Incident #418Uber Locked Indian Drivers out of Accounts Allegedly Due to Facial Recognition Fails	Uber’s facial recognition is locking Indian drivers out of their accounts – technologyreview.com, 2022-12-06 Big problem has come in Uber App! If you cut your hair or shave, then there will be a problem – hindi.news18.com, 2022-12-09
Incident #420Users Easily Bypassed Content Filters of OpenAI’s ChatGPT	ChatGPT proves that AI still has a racism problem – newstatesman.com, 2022-12-10 ChatGPT bot tricked into giving bomb-making instructions, say developers – thetimes.co.uk, 2022-12-17 Testing Ways to Bypass ChatGPT’s Safety Features – lesswrong.com, 2022-12-06 Tweet: @spiantado – twitter.com, 2022-12-05 The Internet’s New Favorite AI Proposes Torturing Iranians and Surveilling Mosques – theintercept.com, 2022-12-08
Incident #421Stable Diffusion Allegedly Used Artists’ Works without Permission for AI Training	Lensa, the AI portrait app, has soared in popularity. But many artists question the ethics of AI art. – nbcnews.com, 2022-12-07
Incident #423Lensa AI’s Produced Unintended Sexually Explicit or Suggestive “Magic Avatars” for Women	‘Magic Avatar’ App Lensa Generated Nudes From My Childhood Photos – wired.com, 2022-12-07 Lensa’s viral AI art creations were bound to hypersexualize users – polygon.com, 2022-12-20 The viral AI avatar app Lensa undressed me—without my consent – technologyreview.com, 2022-12-12
Incident #424Universities’ AI Proctoring Tools Allegedly Failed Canada’s Legal Threshold for Consent	Online proctoring biometrics fails to meet Canada’s legal threshold of consent: report – canadianlawyermag.com, 2022-12-02
Incident #425State Farm Allegedly Discriminated against Black Customers in Claim Payout	New Suit Uses Data to Back Racial Bias Claims Against State Farm – nytimes.com, 2022-12-14 State Farm accused of making it harder for Black customers to get payouts – courthousenews.com, 2022-12-14
Incident #427Cruise Taxis’ Sudden Braking Allegedly Put People at Risk	G.M. Unit’s Self-Driving Taxis Are Subject of U.S. Safety Investigation – nytimes.com, 2022-12-16 ODI Resume – static.nhtsa.gov, 2022-12-12 Do Cruise’s AVs brake-check other drivers? Feds are investigating – arstechnica.com, 2022-12-16
Incident #431Robbers Accessed Drugged Gay Men’s Bank Accounts Using Their Phones’ Facial Recognition	Druggings, Deaths and Robberies Put New York’s Gay Community on Edge – nytimes.com, 2022-12-03 Gay M​​en in NYC targeted by Serial Murder or Hate Group – pride.com, 2022-12-05
Incident #434Sudden Braking by Tesla Autopilot Allegedly Caused Eight-Car Collision	Tesla ‘full self-driving’ triggered an eight-car crash, a driver tells police – cnn.com, 2022-12-21 Tesla behind eight-vehicle crash was in ‘full self-driving’ mode, says driver – theguardian.com, 2022-12-22 Tesla driver in multi-car crash told police self-driving software malfunctioned – reuters.com, 2022-12-22

🗄 From the Archives

Every edition, we feature one or more historical incidents that we find thought-provoking.

Southwest Airlines’ mass flight cancellations highlighted how extreme weather can disrupt the operations of intelligent systems. The following examples from the AI Incident Database show that this is not the first time weather has interfered with intelligent systems.

In late December 2021, Google Maps directed drivers in the Sierra Nevada area to closed-off and even dangerous roads amid record-setting snowfall (Incident #155), while in February 2019 Apple Maps allegedly directed a couple on a ski trip in the mountains toward an unconventional route where the drivers found themselves stuck on an unpaved road in the snow (Incident #288). Self driving cars have also had difficulties navigating snow, as demonstrated in February 2016 when one of Volvo’s autonomous vehicles experienced sensor blindness raising concerns about future issues for self driving cars in snowy regions (Incident #70).

History has shown that rare and unpredictable circumstances challenge intelligent systems, and the weather will continue to surprise humans and break their computer systems.

👇 Diving Deeper

• All new incidents added to the database in the last month, grouped by topic:

  • Privacy & surveillance: #424, #425, #433
  • Facial recognition: #412, #418, #430, #431
  • Bias & discrimination: #429
  • Social media: #414, #415, #416, #417, #419
  • GPT: #413, #420
  • Impactful Errors: #428, #432
  • Autonomous vehicles: #426, #427, #434
  • AI generated images and deepfakes: #421, #422, #423

• Explore clusters of similar incidents in Spatial Visualization

• Check out Table View for a complete view of all incidents

• Learn about alleged developers, deployers, and harmed parties in Entities Page

---

🦾 Support our Efforts

Still reading? Help us change the world for the better!

1. Share this newsletter on LinkedIn, Twitter, and Facebook
2. Submit incidents to the database
3. Contribute to the database’s functionality