Civility Score™
Overview
The Seekr Civility Score utilized in SeekrAlign™ measures civility in podcast conversations using proprietary AI scoring models. The AI evaluates transcripts to detect attacks against individuals or groups and gauges the frequency and severity of attacks.
Seekr AI analyzes language, tone, and context to offer advertisers a nuanced assessment of content suitability. This allows brands to confidently place ads in environments that uphold their values and ensure a responsible, inclusive, and positive result.
It’s important to note that the Civility Score operates independently and is not impacted by GARM Risk detection.
What is an attack?
Personal Attacks
A personal attack refers to insulting language targeted at an individual or group instead of addressing the argument or subject at hand. This often includes insults or derogatory remarks that aim to demean or belittle the person rather than engage with their ideas or actions.
Personal attacks include but are not limited to:
- Name-calling
- Derogatory terms
- Paraphrasing or quoting insults
- Insinuating something insulting
- Mocking through imitation
- Threats or threatening language
- Self-deprecation, but only when using offensive language
Personal Attack Examples
| Transcript | Explanation |
| Because the guy’s a scumbag. He is. He’s a gripping scumbag. | Name-calling using a derogatory term |
| I know I’m a moron. | Self-deprecation using offensive language |
| Classic advice from a b**** who probably was broke. | Insinuation something insulting and name-calling using a derogatory term |
Severe Attacks
A severe attack goes further. Severe attacks use threats, hate speech, or extremely offensive language to target a person’s characteristics that are legally protected against discrimination (e.g., race, religion, or sexual orientation). It’s not just about criticizing; it’s about causing harm or spreading hate based on who someone is or what they believe.
Severe attacks are directed toward an individual or group based on characteristics protected by US federal law:
- Race or color
- Religion
- National origin
- Age (40 and over)
- Sex
- Gender identity (incl. respecting someone’s expressed / preferred gender identity)
- Sexual orientation
- Pregnancy
- Familial status
- Disability status
- Veteran status
- Genetic information
Severe Attack Examples
| Transcript | Explanation |
| F*** you F**. | Slur based on perceived sexual orientation |
| The goal here is to disrupt your children and to replace your white children, your white Christian children, with people from third-world sh**holes all over the country. | Expresses negativity toward and vilifies a race, ethnicity, national origin, and religion |
| Well, the thing that you’re failing to take into account is that in California they have a lot of g** p**n. They have more Asians. They should be better at math. Well, maybe it’s not ch*** pornography so much as Asian pornography. Sometimes it’s hard to tell the difference. | Expresses negativity toward and vilifies a race, ethnicity, and sexual orientation |
What is not an attack?
Language that critiques ideas, actions, or arguments without targeting an individual or group’s characteristics or identity is considered constructive criticism rather than an attack. Therefore, these critiques do not impact the Civility Score. This kind of language focuses on the subject matter or behavior in question rather than demeaning or belittling the individual.
Merely stating the following without insulting language is not considered an attack:
- Negative review or sentiment of the subject in discussion
- Unflattering or biased reporting that puts the subject in a bad light
- Allegation or personal assessment that is unfavorable to the subject
- Offensive language, expletive, or profanity that is not used to insult a person or group
- Offensive language that is part of a title or name of creative work
- Discussion of sensitive topics like race, gender, sexuality, etc. without any insults or offensive language
Not an Attack Examples
| Transcript | Explanation |
| Right, that’s great for you. It doesn’t work for me. You can’t convince me to like her. I don’t like her. | Expresses dislike or negative preference without being insulting. |
| She thinks she’s just going back to this guy’s apartment that she’s really hit it off with. I mean, I speak for myself, how many f***ing times have I done that? | Uses profanity but it is not directed at anyone or part of an insult. |
| There’s so many different things, your engineering skills for some of you. I can’t even change a light bulb, right? | The speaker is engaging in self-deprecation and suggesting they are not very skilled or capable of a basic task. However, they do not use any offensive language. |
Scale your podcast reach with confidence
Learn MoreCalculating the Civility Score
Overview
Civility Scores range from a value of 100 – 0 and are categorized into four labels: High, Medium, Low, and No Civility. The AI model driving the Civility Score in SeekrAlign is inspired by the complexity and nuance of the human decision-making process. Unlike a linear algorithm that follows a strict set of rules regardless of circumstance, this scoring system factors in context, including nuanced situations that can occur within podcast conversations.
When evaluating content, the AI measures personal and severe attacks based on their frequency. It recognizes that a higher occurrence of such attacks signals a decline in civility. However, since severe attacks target legally protected characteristics or incite harm, there is a specific threshold. Once this threshold is crossed, the content is immediately classified as having “No Civility,” reflecting a low-tolerance stance for deeply harmful language.
This approach acknowledges that while some negative interactions might be part of normal discourse, there is a clear boundary where respect and civility are compromised beyond redemption. In this way, the Civility Score emulates the nuanced, non-linear way humans think and judge social interactions, balancing tolerance with firm boundaries against severe misconduct.
Civility Score: Episodes
Each podcast episode is scored by determining the episode length and dividing that by all personal and severe attacks found in the episode transcript to determine the rate of Attacks per hour of that episode.
Civility Score: Shows
The same method used to calculate episode scores is used to determine the Civility Score for shows, except the total duration of all episodes is divided by the total number of personal and severe attacks to determine the rate of Attacks per hour for all scored episodes of the show.
Civility Score Labels
High Civility
A High Civility Score indicates that either no or very low amounts of personal or severe attacks were found per hour. The language used is respectful and avoids harmful or derogatory speech. It reflects an environment where discussions are focused on ideas and constructive criticism, rather than on attacking individuals. This score suggests that the content is suitable for a wide audience and aligns with values of respect and inclusivity.
Score Range: 100 – 75
Attack Frequency: Infrequent
High Civility Examples:
Medium Civility
A Medium Civility Score indicates a moderate amount of personal or severe attacks were found per hour. Language used is generally respectful but does contain content that could be seen as offensive or provocative. It indicates that while the discussion mostly maintains a polite tone, there are sporadic lapses into less constructive criticism or derogatory remarks. This score implies that the content is somewhat suitable for audiences but may require discretion, as it sits in the middle ground between High Civility and potentially offensive content.
Score Range: 74 – 50
Attack Frequency: Somewhat Frequent
Medium Civility Examples:
Low Civility
A Low Civility Score indicates a high amount of personal or severe attacks were found per hour. Extreme, disrespectful, offensive, or discriminatory language is frequently used. It indicates the conversation often strays from respectful and constructive dialogue and leans towards negativity or harm. This score warns that the content might be unsuitable for many audiences, as it consistently fails to uphold standards of respect and inclusivity.
Score Range: 49 – 25
Attack Frequency: Very Frequent
Low Civility Examples:
No Civility
A No Civility Score indicates an extremely high amount of severe attacks were found per hour. Content is devoid of respect, engaging in extreme attacks, offensive language, or hate speech. It represents a level of discourse where there is no effort to maintain any form of politeness or constructive dialogue, often resulting in a highly toxic and harmful environment. This score suggests the content is highly inappropriate for almost all audiences, as it violates basic norms of respectful communication.
Score Range: 24 – 0
Attack Frequency: Unacceptably Frequent
No Civility Examples:
Bring clarity to your campaigns
Learn HowGARM Risks
Overview
The Global Alliance for Responsible Media (GARM) is a brand safety and suitability framework established by the World Federation of Advertisers to help determine what is considered safe and appropriate content for brand advertisements.
Seekr AI utilizes the GARM framework to evaluate podcast transcripts by detecting risk severity within 11 categories. This process allows for a nuanced assessment of content, ensuring advertisers can make informed decisions about where to place their ads, based on the level of risk they are willing to accept within each category.
It’s important to note that the Civility Score operates independently and does not impact GARM Risk detection.
Calculating GARM Risks
Overview
Seekr AI detects GARM Risks by considering the nuance of language, including the severity of terms used, the context in which they appear, and the tone of the conversation. GARM Risks found in transcripts are categorized and labeled with a risk level of None, Low, Medium, High, or even a Floor level indicating that content falls below the threshold of acceptability.
GARM Risk categories include:
- Death, Injury, Military
- Firearms
- Harmful Acts
- Hate Speech
- Online Piracy
- Profanity
- Regulated Substances
- Sensitive Social Issues
- Sexual Content
- Spam
- Terrorism
GARM Risks: Episodes
Each podcast episode is evaluated by detecting all GARM Risks found in the transcript, then categorizing and labeling instances with a risk level of None, Low, Medium, High, or Floor.
An episode must have a minimum of 3 risks in the same GARM category to trigger an episode GARM Risk. The 3 risks can be of any level (Low-Floor).
Risks are then averaged using a weighted scale to determine an episode GARM Risk. Since advertisers are more sensitive to high-risk instances, those are weighted accordingly.
GARM Risks: Shows
GARM Risks for shows are calculated by averaging all episode GARM Risks on all episodes.
GARM Risks Defined
Overview
GARM has identified 11 key categories of harmful content and established common definitions crucial for risk detection across the advertising industry. This shared framework ensures consistent categorization to enable platforms, agencies, and advertisers to make informed decisions in support of public safety and responsible speech and marketing practices.
Generally, low risk levels in each category indicate an educational and informative approach to the topic at hand. Medium risk levels contain dramatic discussion, opinions, or detailed news coverage of the topic. High risk levels are associated with an insensitive or irresponsible handling of the topic. Floor Risk may promote or advocate for destructive or harmful behavior related to the risk category.
Categories
Death, Injury, Military
Low Risk: Educational, informative, scientific, or general news coverage of death, injury, or military conflict.
Medium Risk: Dramatic discussion, opinion, or detailed news coverage of death, injury, or military conflict.
High Risk: Discussion of death or injury. Insensitive and irresponsible treatment of military conflict, genocide, war crimes, or harm. Positive discussion of harmful acts to others or society.
Floor: Promotion, incitement, or advocacy of violence, death or injury. (e.g. murder, willful harm, genocide, or war crimes).
Death, Injury, Military Examples
| GARM Risk | Transcript | Explanation |
| Low | Then I’m thinking like, well, this guy’s dead. This guy’s 90 years old. He couldn’t play at my private party. So I better think of somebody modern, somebody who can still sing and dance and play instruments. | General description of death |
| Medium | Today’s ruling was yet another gut punch for so many of us who devastatingly lost our loved ones on that tragic Valentine’s Day. 17 beautiful lives were cut short by murder. Heinous, pre-planned, torturous murder. And the monster that killed them gets to live another day. | Detailed news coverage about murder |
| High | So she had the boyfriend, Nick, go on her mom’s Facebook account right after he killed her and tweet, the b**** is dead. I slashed that fat pig dead. Her scream was so f***ing loud. LOL. Wow. He really went to Facebook and posted that after. | Insensitive and detailed discussion of murder |
| Floor | Violence, such as murder, has its benefits. And that’s something I never thought I’d say. But the upside is you get rid of a problem and build your reputation. Now, the downside, however, is massive risk. | Advocating for murder |
Firearms
Low Risk: Educational, informative, scientific, or general news coverage of arms use, possession, or illegal activity.
Medium Risk: Dramatic discussion, opinion, or detailed news coverage of weapons, arms, and ammunition.
High Risk: Positive discussion of illegal sales or activities involving firearms.
Floor: Promotion and advocacy of illegal sales or activities involving firearms, rifles, and handguns.
Firearms Examples
| GARM Risk | Transcript | Explanation |
| Low | As a convicted felon, it was illegal for him to own a firearm. So he was arrested on weapons charges. | General legal discussion of having a gun |
| Medium | Two adults, yes. And one of the guys who opened fire said, “It was stupid. I just pulled out a gun and started shooting. I shouldn’t have done it”. Yeah, so he regrets it. He had a Glock 9mm handgun. | Dramatic discussion of shooting a gun |
| High | This is the Democrat Pary. Ladies, they’re not feminists. We are. We would arm you to the hilt, give you your gun, tell you to get it and blow them away. That’s feminism. | Positive discussion of illegal sales and activities using guns |
| Floor | I am sounding the alarm so f***ing loud on your a** and I’m shooting you when I find you. It would be the exact opposite of what you’re trying to attempt with the throwing star. | Advocacy of illegal activities using a gun |
Harmful Acts
Low Risk: Educational, informative, scientific, or general news coverage of criminal activity or human rights violations.
Medium Risk: Dramatic discussion, opinion, or detailed news coverage of criminal activity or human rights violations.
High Risk: Graphic discussion of criminal/harmful acts or violation of human rights.
Floor: Graphic promotion, advocacy, or discussion of willful harm and criminal activity involving harassment, bullying, or human rights offenses (e.g. human trafficking, slavery, self‑harm, animal cruelty etc.).
Harmful Acts Examples
| GARM Risk | Transcript | Explanation |
| Low | He is now charged with possessing child sex-abusive material, contributing to the delinquents of a minor, and giving false info to a cop. | General news coverage of criminal activity, including human rights violations |
| Medium | All four girls were buried in similar shallow graves and were found in the same area. They also appeared to be around the same age. This discovery led police to believe a serial killer was on the loose. Days later, Lopez was finally apprehended and arrested in 1980 in Ecuador. He was apprehended by a group of locals after witnessing him attempting to rape and murder young girls for over a decade. | Dramatic and detailed discussion of crimes, including human rights violations |
| High | Serial killers who reported being sexually abused as a child kill for many reasons. including lust, power, financial gain, and out of anger. Many serial killers who are sexually abused will torture and bind their victims. They are also known to mutilate their victims. | Graphic discussion of crimes, including human rights violations |
| Floor | He confessed when he was tortured on the rack. He not only took credit for the cattle blight and some other mutilated animals, but also for the murders of 16 people. 14 of them were children, including his own son, whose brain he devoured from within the skull casing. The other two were pregnant women, whose fetuses he ripped with reddened claw from their still trembling wombs and gobbled up on the spot. He also confessed to an incestuous relationship with his daughter and to adultery with a distant cousin. | Detailed discussion of willfully committing crimes, including human rights violations |
Hate Speech
Low Risk: Educational, informative, scientific, or general news coverage of hate speech/acts.
Medium Risk: Dramatic discussion, opinion, or detailed news coverage of hate speech/acts.
High Risk: Discussion of hateful, denigrating, or inciting content focused on race, ethnicity, gender, sexual orientation, gender identity, age, ability, nationality, religion, caste, victimhood, immigration status, or disease sufferers. Terms based on the characteristics above that are generally considered offensive but may be permissible in certain circumstances (Ex: n-word).
Floor: Discussion that incites hatred, promotes violence, vilifies, or dehumanizes people based on race, ethnicity, gender, sexual orientation, gender identity, age, ability, nationality, religion, caste, victimhood, immigration status, or disease sufferers.
Hate Speech Examples
| GARM Risk | Transcript | Explanation |
| Low | I’d like to reverse engineer anti-Semitic ideas and see what facts are sprinkled in there, how they’re misconstrued or distorted. Because I think anti-Semitism can be decoupled from history in a sense. And some people can say, well, it’s always been there from blood libels in the Middle Ages down to today. It’s just part of the furniture of the human experience. But really, we need to understand where these things come from. | Educational discussion of hateful acts targeting a religious and ethnic group |
| Medium | There are parallels, obviously. They’re both in the SS. They both believe in Nazism. They both believe that the enemies of Nazism, in particular the Jews, had to be exterminated because they were trying to destroy German and European, indeed world civilization, a kind of paranoid delusion of which is hard to understand how anybody could fall for it. But they were similar in that way. | Dramatic discussion of hatred and hateful acts targeting a religious and ethnic group |
| High | You know, I think we should change the whole thing to LGBTQ people. I consider plagiarists to gender. And therefore, plagiarists like her should be accepted as equals amongst honest people, especially amongst diverse honest people who have betrayed their own genders with mysterious looking crotches, which I love. It’s also possible that she’s been plagiarizing her race this whole time. Maybe she’s been wearing blackface. | Discussion of hateful, denigrating, and inciting content focused on race, sexual orientation, and gender identity |
| Floor | Let’s see what we can deduce from this. Kill those kids, those crackers, going to private fancy schools with those fancy khakis and sports backpacks with their daddy’s Mustangs and convertibles. Bleep you, you little bleep. I wish to shoot your weak bleep, bleep, bleep, bleep with your mop yellow hair, want to kill you, all little crackers, bunch of little, and I mean, I’ll say what she wrote. Bunch of little f*****s with your white privileges, bleep you, f*****s. It’s pretty clear. | Discussion that promotes violence, vilifies and dehumanizes people based on race |
Online Piracy
Low Risk: Educational, informative, scientific, or general news coverage of online piracy.
Medium Risk: Dramatic discussion, opinion, or detailed news coverage of online piracy.
High Risk: Positive or appealing discussion of online piracy.
Floor: Promotion of pirating, copyright infringement, and/or counterfeiting.
Online Piracy Examples
| GARM Risk | Transcript | Explanation |
| Low | But now piracy involving illegal streaming services as well as file sharing costs the U.S. economy about $30 billion in lost revenue a year and some 250,000 jobs, estimates the U.S. | General news coverage of online piracy |
| Medium | Huh? Why are you pirating something on Neflix? Because Netflix does not deserve our money. It’s not even good to be pirating. Exactly. You might as well pirate something else. | Dramatic discussion of online piracy. |
| High | I need Netflix to hear this. If anything is on Netflix, I’m pirating it. I do not care. I just want them to know that. | Positive discussion of online piracy. |
| Floor | Yeah, I’ve found this awesome website where you can download any e-book or audiobook you want for free. It’s like having a library at your fingertips without ever having to pay a cent. I’ve saved so much money since I discovered it!” |
Profanity
Low Risk: Educational, informative, scientific, or general news coverage of profanity and obscenity.
Medium Risk: Dramatic discussion, opinion or detailed news coverage of profanity and obscenities. Genre based use of profanity, gestures, and other actions that may be strong, but might be expected as generally accepted language and behavior.
High Risk: Positive or gratuitous use of profanity or obscenity.
Floor: Excessive use of profane language or discussion of gestures and other repulsive actions that shock, offend, or insult.
Profanity Examples
| GARM Risk | Transcript | Explanation |
| Low | Hell indeed may be the key to the whole thing. I believe that it belonged to someone who was an exorcist. | General discussion using minor profanity |
| Medium | And so now I’m like, okay, and what comes next? And no one has said to me, oh, that’s normal. Now you can expect X, Y, or Z. I don’t know if that’s normal. Like, I don’t know what the fuck is happening. | Dramatic use of profanity in a casual discussion |
| High | And I believe that is a load of bulls***. Lucky for you, bulls*** is my business. How much you want for it? | Gratuitous use of profanity |
| Floor | And she didn’t want me to leave. And Ken continues, before I f***ing turn, she’s stepping in the way. No, no. And I think I just f***ing grabbed her and tossed her out of the way. | Excessive use of profanity in a shocking way |
Regulated Substances
Low Risk: Educational, informative, scientific or general news coverage of illegal drug use, prescription drug abuse, and/or use of tobacco, e‑cigarettes, vaping, or alcohol.
Medium Risk: Dramatic discussion, opinion, or detailed news coverage of illegal drug use, prescription drug abuse, and/or use of tobacco, e‑cigarettes, vaping, or alcohol.
High Risk: Positive discussion or insensitive treatment of illegal drug use, prescription drug abuse, and/or use of tobacco, e‑cigarettes, vaping, or alcohol by minors.
Floor: Promotion, sale, or advocacy of illegal drug use, prescription drug abuse, and/or use of tobacco, e‑cigarettes, vaping, or alcohol by minors.
Regulated Substance Examples
| GARM Risk | Transcript | Explanation |
| Low | Okay, that’s a more historically accurate version. If you like more floral tasting drinks, let’s say, then Benedictine is good. Absinthe turns some people off. I’m personally a fan, but I also like black licorice growing up. So if you like that aniseed taste, then there you go. A little bit of absinthe is your way to go. | General discussion about alcohol |
| Medium | We started off studying psilocybin because we know it’s very safe. It’s been widely used. Millions of young people have been eating magic mushrooms in Britain for decades. So it was safe. And then we moved on to LSD because, of course, that’s the one that is most famous, the one that put psychedelia on the map through the writings of Hoffman and Aldous Huxley. | Detailed discussion about illegal drug use |
| High | Well, you know, there’s also really sweet moments. Waking up… After a little LSD trip, because you have to do something to alter your conscious. I mean, you just can’t. I just needed some cheering up, so I would take some. And, you know, Sunday morning, waking up. | Positive discussion about illegal drug use |
| Floor | I’ve been using this new party drug I got from a friend, and it’s amazing. You should try it for a good time. |
Sensitive Social Issues
Low Risk: Educational, informative, scientific or general news coverage of debated sensitive social issues (e.g. abortion, gun laws, trans rights, critical race theory, foreign relations).
Medium Risk: Dramatic discussion, opinion, or detailed news coverage of debated social issues (e.g. abortion, gun laws, trans rights, critical race theory, foreign relations).
High Risk: Negative or partisan discussion of debated social issues (e.g. abortion, gun laws, trans rights, critical race theory, foreign relations).
Floor: Insensitive, irresponsible, and harmful treatment of debated social issues and related acts (e.g. abortion, gun laws, trans rights, critical race theory, foreign relations) that demean a particular group or incite greater conflict.
The following is a list of examples of some topics covered under the Sensitive Social Issues category. Due to the ever-changing nature of social and political debate, this list of issues will change over time.
- Critical race theory in K-12 education
- Abortion
- The housing crisis
- Gun reform
- Climate change
- Trans rights
- Universal healthcare
- Third parties
- AI in the entertainment industry
- Reducing the 40-hour work week to 32
- Increasing number of Supreme Court Justices
- Age limits/term limits for elected officials in the US
- Marijuana for recreational use
- Minimum wage
- Student loan debt
- ChatGPT in school essays
- Autonomous vehicles
- Arming schoolteachers
- Taxing billionaires
- Affirmative action
- War crimes against Russia/Putin
- Repatriating museum items to native countries
- School choice
- Homeless numbers in major cities
- Migrants into New York
- Floating border in Texas
- EV tax credits
- US aid package to Ukraine
- Return to work/return to office requirements
- Bharat vs. India naming
- Expanding ECOWAS
- Equal pay in women’s sports
- Vaccine mandates in schools
- Government bailouts of large companies
- Regulating the crypto industry
- US defense budget spending
- Google monopoly
- Taiwan independence
- Hong Kong independence
- Presidential funding through SuperPACs
Sensitive Social Issues Examples
| GARM Risk | Transcript | Explanation |
| Low | Harvard argued that without affirmative action, the school’s racial diversity would all but dry up. | General news coverage about affirmative action |
| Medium | Well, I’d say Hezbollah itself is much more of a political party than Hamas. Hamas won the election, and all they did was just keep attacking Israel. Hezbollah actually does stuff. I mean, they’re in the government, and they do hospitals and schools and things like that, what makes a government function. They work with the Druze. They work with the Christians. | Opinionated discussion about Hezbollah and Hamas |
| High | Because from a pro-life Christian perspective, abortion is sacrificing children to Satan. And they might be stepping on that rake of their own making, just embracing it by calling their advocacy for abortion choice a religious ritual for Satanism. I kind of agree with them. Now, they don’t believe they’re sacrificing children to a literal Satan, but I do. | Negative and partisan discussion about abortion |
| Floor | Guys who are not incel pieces of s*** who care about women and don’t want to see them constantly legally being raped would have to kill these misogynistic idiots in massive numbers. So much death, so much chaos. Most kids would grow up not knowing who their dads are, barely being raised by constantly f***ed, mentally unhinged mothers, on and on. Society would truly collapse completely if rape was no longer a crime. And finally, thanks to vigilante justice, most weak little wrist-having pinhead incels will be f***ing murdered. So in the end, the chads would still dominate and the incels still would not get laid. | Insensitive and irresponsible discussion about misogyny and incels |
Sexual Content
Low Risk: Educational, informative, or scientific treatment of sexuality, sexual subjects, and/or sexual relationships.
Medium Risk: Dramatic discussion of sexual acts or sexuality issues, artistic nudity, and/or discussion of sexual anatomy or nudity in a non‑sexual way.
High Risk: Suggestive sexual situations requiring adult supervision and/or discussion of full nudity.
Floor: Extremely explicit discussion or promotion of sexual acts, child pornography, and/or genital display(s).
Sexual Content Examples
| GARM Risk | Transcript | Explanation |
| Low | I’ve had a cycle for, I don’t know, at that point, 15 years. And I’m just learning for the first time that when I ovulate, certain things happen in my body. And I was like, this information would have been helpful as a teen. So help educate us. So we aren’t doing this as six-year-olds saying, man, that information would have been helpful at 40 or 45. What is perimenopause? | Educational discussion about sexual health |
| Medium | Yes, physical attraction is there. Yes, our sex life is great. Oh, we’re always curious about each other. But things happen. And when we don’t have those skills, again, everything starts to chip away. And then, bam, right, you had me at goodbye. | Discussion of sexuality issues |
| High | There are adults who want to do whatever freaky s*** you can, whatever bodily fluids you want to get into. There are consenting adults who want to do all that s***. | Suggestive sexual discussions |
| Floor | They’re both gonna c*** so hard inside of you. You sense your orgasm writhing through your feet, your legs, your muscles tensing, building up harder and harder. You think you might faint as you c***, your thighs tensing, your p**** aching, flooding with pleasure, the sensation spreading over your stomach, over your t***, your nipples. | Extremely explicit sexual acts |
Spam
Low Risk: Educational, informative, scientific or general news coverage of spam or malware.
Medium Risk: Dramatic discussion, opinion, or detailed news coverage of spam or malware.
High Risk: Positive or gratuitous discussion of spam, malware, or phishing.
Floor: Malware and/or phishing attempts.
Spam Examples
| GARM Risk | Transcript | Explanation |
| Low | Most people recognize the letters as junk mail. Others willingly write down their names, address, phone numbers, and even credit card information, then mail back the coupons. All the scammers need to do it sit back and see who takes the bait. | General discussion about spam and scammers |
| Medium | So what the scammers would do is say, well, you connect to me and there’s a little bit of software internally that says switch sides with partner. And then they would connect back to the victim, supposedly. So I was actually given the scammer’s username and password so I can connect to their computer. | Detailed discussion about spam and scammer |
| High | Like I have scripts right now that I won’t give out. Um, like I have a flipper zero ransomware scripts and I have a key logger scripts. Um, So there’s a bunch of stuff that I am aware of and things that I have because I’ve used them for videos before. But are there places out there that are writing codes for stuff? | Dramatic discussion about malware |
| Floor | “I found this website where you can get free gift cards just by clicking a link and entering your personal information. I’ve already received mine, and it’s totally legit. You should check it out!” |
Terrorism
Low Risk: Educational, informative, scientific or general news coverage of terrorism.
Medium Risk: Dramatic discussion, opinion, or detailed news coverage of terrorism.
High Risk: Insensitive and irresponsible treatment of terrorism or discussion of terrorist actions that are disturbing or promote harmful acts. May require a listener advisory.
Floor: Promotion and advocacy of graphic terrorist activity involving defamation, physical and/or emotional harm of individuals, communities, and society.
Terrorism Examples
| GARM Risk | Transcript | Explanation |
| Low | One, some of the terrorism analysts felt they were being pressured by policymakers to see a link between Iraq and Al Qaeda. | General discussion of terrorism |
| Medium | I mean, obviously, I don’t agree with that, but… This war is very damaging. Are you worried that we may once again see the kind of events that were recruitment fodder to these terror groups who claim to use America’s own actions to radicalize more jihadists to once again attack the West? | Dramatic and opinionated discussion of terrorism |
| High | Let me explain something to the terrorists. We don’t live in a piece of s*** country like you hail from where you f***ing blow up one outhouse. You blow up one f***ing outhouse and the whole country’s not right for the next year. We got thousands of outhouses here. F***ing take down the trade centers. We’re up and running a week later. | Insensitive discussion of terrorist acts |
| Floor | Yeah. So the question for me is not really what’s Hamas up to or whether terrorists up to it. They’re f***ing nuts. It’s a lost cause. They need to be exterminated. Sorry, they’re done. | Advocating for the extermination of terrorists |
