Good afternoon gentlemen (and ladies),

I come from other parts of the internet, and exposing shills, bots, socks, the technology they use and who uses them was a thing I used to do, and I will share my discoveries:

1) Image based captcha don't actually work on modern tech as most bot accounts have smart OCR systems that can detect even 'pretty bad' images and answer the captcha - the one you use for this board wouldn't even stop them (as you've discovered)

2) Text based questions can be answered by one of two systems:

2a) The operator 'adds the answer', or

2b) An AI inference engine tries to 'guess' or 'interpret' the question (EG What is 2 + 2? 4, for example)

3) You can thwart most bots by having a question/answer that changes on a daily basis (with no reuse) on top of a captcha, this can impede users however.

4) Bot operators are called 'handlers', and they can operate between 2 to 100 accounts. The average is between 15 to 30.

5) The software they use is commercially developed, although it appears some are homebrewed applications that run in Python (possibly to pair up with TensorFlow?)

6) Building a forum bot is incredibly easy: Python with mechanize library and BeautifulSoup (you might want to consider fighting fire with fire).

7) Governments, Corporations and Political groups all use bots. A prime example of corporations is the Net Neutrality spam. They use it to astroturf grassroots movements and artificially incite aggression/action/results (think Arab Spring).

8) Organisations confirmed: IDF (both professional and hired students), Mossad, CIA, NSA (Edward Snowden admits as such), parts of MI5, US Airforce, The MoD (specifically: the Army), MI6 (minor role), FBI (usually to honeypot/entrap/snare), GCHQ, FSB and whatever the KGB's replacement is. Democrats are more active with social engineering/bots/socks, Republicans are playing catch up. British politicians have yet to catch on, it seems. Pretty much every corporation does it - usually they send in their employees to write positive blurbs or counter-criticism (look at any critical article about Microsoft, Red Hat, systemd etc).

9) It costs a lot of money, but gets cheaper with automation and AI.

10) Handlers can't be in more than two places at once (mentally speaking: the bots log/alert the replies), and the more people simultaneously engaging them from different places (as their 'differing' accounts), the better as it overwhelms their responses and often they make mistakes.

11) Their bots are often 'tuned' to specific keywords and are capable of bulk-scanning threads (usually the last page to check for responses). If you mess up the keywords and don't mention them directly, the alert system appears to fail - they will get notified by other bot handlers however if it's something they've missed, and the new keyword will be added.

12) Some bot/sock accounts talk amongst themselves. Usually so one sets up, the other introduces information.

13) They will strawman arguments - present a weak front under one account, attack it with another. Don't accept or try to defend weak arguments.

14) The different groups have different agendas, from what I can see:

IDF: Solely concerned with topics that directly relate to israel, berating palestine (treating palestinians as non-human etc), attacking iran, tackling anti-semitism

US agencies: Whatever the current US politics are, always pro-military

UK agencies: Whatever the current UK politics are, always pro-British, sometimes pro-US (they have minor disagreements)

Russia: Usually anti-agency/anti-US/UK establishment, will bring up topics that advantage them (Syria, Palestine, US abuses of power)

Corporations: Anything directly relating to them (usually they target criticism)

Political groups: Shilling in favour of their political agenda/who they want you to vote for/attacking other parties.

I hope my observations are of some use. If I think of more I'll get back.

Thought of more that might be relevant:

1) Advanced software (like in HBGary) will sometimes try to create 'backstories' or 'multiple accounts' across social media outlets - this is to make the poster seem more legitimate than they actually are. The one thing they can't easily create is 'history' (especially once outed). Some bot/sock accounts - in rare exceptions - are 'sleeper accounts', that are created historically some time earlier that might be dragged into bigger fights (so they appear more legitimate to a skeptical audience).

2) They utilise ISPs or systems that give high IP rotation (EG 'mobile', Verizon Business, satellite, etc) - essentially IPs that 'muddy waters' and make it difficult for source attribution. Proxies are less common but do get used. Normal posters tend to have consistent IPs or ISPs in contrast.

3) Handlers will target what are known as 'leaders' in social media circles. These are the people who are influencing the rest (Q an example, but also board owners, admins, that sort of thing - anyone who has a 'big presence'). The goal is to disrupt the leaders in order to disrupt the followers. They may also try to insert 'counter' leaders (who emulate the main leader in some way) who are basically like 'mimicry' style leaders. The counter to this is for everyone to independently think on their own terms rather than a 'leader-follower' network.

4) Government level, and some high level corporation (think military industries type shit) have what's known as 'writing analysis' tools. This allows them to unmask anonymous posters on the internet (kinda like how you identify shills based on how they act/behave - except automated). Ironically this can be used against scripted bots if you have someone smart enough to build your own version.

4a) The analysis tool, from what I can gather works on the following:

Post length (average length of posts).

Unique/key words (British people use different words to Americans, for example, but also overusing specific memes)

Unique phrases/sentences (you might have a common phrase or sign-off)

Spelling error rate (smarter people are more literate, for example)

Other match methodology that I'm not familiar with

5) A huge tip-off for bots is if they operate 'around the clock'. Normal people - if unemployed - can only really work for one 24 hour segment at most (sleep deprivation kicks in), and at normal, 12 hours. Bots often will be up/posting with a four hour gap or less between posts (ID by IP, post content and spacing). Contrast workplace/homeplace IPs (anyone posting from home then work needs at least two distinct ISPs).

6) Another handler tip-off is they will either claim to be a student (in some cases they actually are: israel pays students to shill for israel, public knowledge, look it up) or 'retired' as an explanation for why they aren't working. In some rare cases they will claim to be 'self-employed' but with apparently no work ethic (basically: look for absences that should be there if 'normal').

7) Horse's laugh/ridicule is a common discredit tactic, but can be made to backfire by earmarking them as immature/unable to handle the mature subject material.

8) Shills are, surprisingly, human. Generally, two types of shill: zealots/psychopaths, and punch-clock shills. Zealots (EG IDF) believe in the cause (they genuinely believe you're evil), psychopaths do it for the lulz, neither can be convinced otherwise. Punch-clock shills only do it for pay/lack of alternative. Edward Snowden was a punch-clock shill - he defected. Break the fourth wall, reach out. They know they're deceiving citizens.

Some shills even purposefully subvert their own work, and will make it obvious - on purpose - that they are a shill. They do have 'supervisors' so they can't be subversive all the time. Can't give examples, their 'supervisors' would clamp down on it. Appeal to the handler, not the bot. A 'good' shill is one you can't see.

Homer once said in response to a bad workplace: "You don't quit, you just do your job half-assedly"

Remember, some will have families, so there's no 'exit stage left'. Expose the shills, but try to also appeal to their humanity (if they have any). They will already know they're working for the bad guys on the simple fact they're attacking a forum full of people trying to expose corruption and human trafficking. Consider posting alternative job opportunities for whatever place they're from.

To the shills:

Do what you can to subvert the deception at your place of work. Even occupied France had the French resistance. A timely tip-off hint here or there can make all the difference.