Is LinkedIn facing a trust and ethics crisis with AI?
LinkedIn is rapidly integrating a wide range of GPT-powered features into its platfrom. Sue Parker, owner of DARE Group, can see how the rush could end in disaster.
LinkedIn is the world’s largest professional network, celebrating its 20th anniversary this year – a milestone that has coincided with the explosion of ChatGPT and OpenAI, of which LinkedIn’s owner Microsoft has invested $10 billion.
The platform’s mission statement is: ‘Connect the world’s professionals to make them more productive and successful’.
The core value proposition is for members to engage in genuine conversations, communities and learnings to develop economic opportunities.
The nucleus for success starts with trust across 930 million global and 13 million Australian members. While there are more than 58 million registered companies on LinkedIn, the hub of connection and conversations comes from individual members.
I am a huge advocate of LinkedIn and embrace change and AI. However, recent LinkedIn and external AI developments raise questions of trust and ethics that need scrutiny.
State of play
Trust is defined as a firm belief in the reliability, truth, or ability of someone or something. But social media fares poorly in the trust stakes, with the 2023 Edelman Trust Barometer Report finding it’s the least trusted industry sector at 44%.
Other global surveys rate LinkedIn as the most trusted of all platforms. And like every other social media, the clamber for influence, visibility and revenue brings out the unscrupulous alongside the ethical with fine lines in-between.
Fake profiles, paid followers and engagement and other sly practices by members are equally prevalent on LinkedIn.
Generative AI and ChatGPT will, and is accelerating the scope of manipulation and fake influence.
Content and comment AI manipulation
Content and engagement/comments are key pillars for success but are being compromised with multiple culpabilities.
As expected, LinkedIn is rapidly integrating a wide range of ChatGPT-powered features. The latest, an AI post generator feature, was launched last month by Keren Baruch, Director of Product.
While still in testing and roll out mode, by using a 30-word prompt, members instantly creates AI posts of opinions, advice and learnings. This will encourage misrepresentation of members’ expertise, a decline in authentic content and a maelstrom of duplicated banality.
But more unsettling is the deluge and increasing uptake of generative AI commenting apps. AI talking to AI is a threat for trust, safety and genuine relationships.
Examples of apps include Engage AI; BrandEngine.ai and Tappy. The latter is particularly pernicious with a tone option of ‘let’s argue’.
Many sellers have a LinkedIn presence and flagrantly promote their ‘prohibited’ services, of which LinkedIn seemingly overlook.
The user agreement expressly prohibits the use of third-party software, extensions, bots and browser plug-ins of which generative AI comment apps and tools sit.
The professional community policies further state that members must make an effort to create original content, respond authentically to others’ content and not falsify information about themselves.
Opinions from industry leaders
I asked a few industry leaders who are prolific LinkedIn users to share their thoughts on these AI developments:
Mark Ritson, Founder Mini MBA Marketing: “Stupid, stupid idea. The whole point of LinkedIn is connecting to other professional people and their genuine thoughts and actual situation. Allowing, never mind encouraging users to allow computers to flood the app with artificial, spooky, dumb, expected spam is a massive punch in the company’s own head. Whoever made this call is a moron.”
Tracey Spicer, Author, MAN-MADE – How the bias of the past is being built into the future: “This is a concerning move. ChatGPT has the potential to reduce the burden of menial tasks in the workplace. However, it’s deeply flawed and filled with bias and prejudice. As users, we need to focus our critical thinking skills on de-biasing this tool, as much as possible. However, the tech giants are expecting us to do this for free, as unpaid labour. Big Tech needs to take bias and equity seriously, before unleashing this technology on the general public.”
Tom Goodwin, digital transformation consultant and author: “These days everyone wants to build a personal brand and get heard but nobody wants to spend the time or energy to think, so as such we’ve all manner of content with zero effort required. People fail to realise the only metric that counts isn’t impressions, but the impression your comment leaves. Which tends to be nothing unless a human could be bothered to have an opinion. One can see this as a way to ensure new voices and ideas get spread by those less confident or practiced in writing. Or, one can see it as a way to jam up the feed with even more lowest common denominator averageness to wade through.”
Alice Almeida, MD, Almedia Insights: “I am not comfortable with AI/ChatGPT being used for content creation or generating comments on LinkedIn without a disclaimer of use. As the human element is part of the LinkedIn experience, I want to know that the person who wrote a post or commented on mine is just that – a person. I go to LinkedIn to learn from people’s experience and knowledge. I want to know that challenging my opinion has come from their own mind and not from something they entered in ChatGPT. Keep it genuine and authentic, or else future meetings in person may be a major disappointment to the other person.”
Legal considerations
I asked Dr Fabian Horton, lawyer and chair of the Australasian Cyber Law Institute to provide a legal lens. He shares Tracey Spicer’s concerns of the embedded biases in AI generated content.
He advised there are many matters to consider including data protection and privacy, intellectual property, issues stemming from discriminatory or biased content and particularly, misleading or false information.
The Australian Consumer Law protects against conduct that is misleading or deceptive or likely to mislead or deceive. Untruthful or inflated claims about products or services could run afoul of commercial or consumer laws.
Dr Horton cautions people to review and ensure AI generated content does not lead to legal ramifications in the areas of defamation, bullying or hate speech.
Whilst LinkedIn supplies automated authoring tools, users are still liable for the content that is posted under their profile. Users should familiarise themselves with LinkedIn’s terms of service and the relevant laws in their jurisdiction to ensure they are not breaching contractual obligations or any civil or criminal laws.
Final wrap
At every touchpoint humans drive decisions, their personal brands and ethics compass.
In a world of relentless change and turmoil, trust and ethics should be the currency of professional sustainability and success, both on and off LinkedIn.
Sue Parker is the owner of DARE Group Australia – a communications, LinkedIn and career branding specialist.
At Alice: “I want to know that challenging my opinion has come from their own mind and not from something they entered in ChatGPT”
What difference does it make, if the challenge to your opinion is logical and reasonable? At least AI will present a cohesive argument, stripped of any personal bias.
User ID not verified.